Selection and Cross-breeding in Relation

to the Inheritance of Coat-pigments

and Coat-patterns in Rats and

Guinea-pigs

BY

HANSFORD MAcCURDY AND W. E. CASTLE

WASHINGTON, D. C. !

PUBLISHED BY THE CARNEGIE INSTITUTION OF WASHINGTON
MAY, 1907

Selection and Cross-breeding in Relation

to the Inheritance of Coat-pigments

and Coat-patterns in Rats and

Guinea-pigs

BY

HANSFORD MAcCURDY AND W. E. CASTLE

WASHINGTON, D. C. I

PUBLISHED BY THE CARNEGIE INSTITUTION OF WASHINGTON
MAY, 1907

CARNEGIE INSTITUTION OF WASHINGTON PUBLICATION No. 70.

PAPERS OF THE STATION FOR EXPERIMENTAL EVOLUTION, No. 8.

CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE MUSEUM OF

COMPARATIVE ZOOLOGY AT HARVARD COLLEGE.

E. L. MARK, DIRECTOR. No. 188.

PRESS OF GIBSON BROS.
WASHINGTON, D. C.

CONTENTS.

Page.

Continuous versus discontinuous variations as factors in evolution .... i

Variations of rats in coat-color and coat-pattern 4

Mendelian inheritance of coat-colors and coat-patterns in rats 7

Albinism recessive in relation to all types of pigmentation 7

Latent pigment characters and coat-patterns 8

Modification of hooded pattern by crossing with Irish 1 1

Modification of hooded pattern by selection 14

Selection for reduced stripe 14

Selection for stripe of increased size 17

Individual spots of guinea-pigs not unit characters 1 8

Color-patterns selected 1 8

Dutch-marked series (Series D) 20

Head -spot series (Series //) 23

Statistical analysis of the data for Series D and // 25

Nose-spot series (Series N) 26

Conclusions 32

Tables 34

Bibliography 50

in

SELECTION AND CROSS-BREEDING IN RELATION TO
THE INHERITANCE OF COAT-PIGMENTS AND COAT-
PATTERNS IN RATS AND GUINEA-PIGS.

BY HANSFORD MACCURDY AND W. E.

CONTINUOUS VERSUS DISCONTINUOUS VARIATIONS AS
FACTORS IN EVOLUTION.

It is generally agreed that the course of evolution is largely influenced
by two factors, variation and heredity; but opinions differ as to what sorts
of variation have evolutionary significance and as to the manner of their
inheritance.

It has been recognized by several investigators that variations are of
two distinct sorts. Bateson has called these two sorts of variation contin-
uous and discontinuous ; more recently De Vries has called them fluctuations
and mutations, respectively.

By continuous variation (or fluctuation) we understand ordinary indi-
vidual variation within a species. The individuals differ among themselves
in size, color, and other structural features. By examining a considerable
number of them we can form an idea of what is the commonest (or modal)
condition as regards each structural feature ; and likewise what is the average
(or mean) condition.

Usually, but not always, the modal and mean conditions are approximately
the same, and any other condition is the less frequent in occurrence, the
greater its deviation from them. It follows that the most extreme condi-
tion observed is connected with the most usual (or modal) condition by
an unbroken series of intermediate conditions, and we may call the series
as a whole "continuous." The distribution of the individuals in such a
series is governed by the laws of "chance," and may be successfully
analyzed by statistical methods.

We commonly think of a "chance" result as something entirely beyond
the control of law, but in reality such is not the case. Nothing is beyond
the control of law. If a blindfolded person puts his hand into an urn con-
taining a mixture of black and of white balls, it is a matter of chance whether
he grasps a black or a white ball; but if he repeats the operation a con-
siderable number of times, it is perfectly certain that he will draw balls of

i

2 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

both sorts in approximately the same proportions in which they occur in
the jar. The result is a "chance" one, but controlled by a perfectly definite
mathematical law.

A "chance result" has been aptly defined as the result of a number of
causes acting independently of each other. If this is a valid definition,
then a continuous series of variations is due to no single cause but to several
mutually independent ones. Some of the causes may be external in origin,
others internal; some temporary in their action, others permanent. It
should not surprise us, therefore, to find that continuous variations differ
greatly in the degree of their inheritance. De Vries, indeed, has maintained
that they are not inherited at all, except temporarily; that selection of
abmodal variations from a continuous series is unable permanently to modify
a race ; that the modifications will persist only so long as selection continues,
but will speedily disappear when selection is arrested. This conclusion,
however, seems to us altogether too sweeping. A priori there is no reason
to suppose that all the causes operative to produce continuous variation
are external in origin and temporary in action, as De Vries's conclusion
would seem to imply. If there are in operation, in the production of a
continuous series of variations, causes internal in origin, resident in the
constitution of the germinal substance, so much of the result as is due to
those causes should be inherited and so should be permanent. De Vries, we
believe, has overlooked this factor entering into the problem. He has
assumed that all the causes of continuous variation ("fluctuations") are
either external in origin or due to conditions of the germinal substance
purely temporary. He holds, we believe rightly, that all inheritance is due
to germinal modification ; but assumes, we believe without sufficient warrant,
that permanent germinal modification is not a factor in the production of
fluctuations.

Another category of variations, discontinuous variations (which include
the mutations of De Vries), is considered by Bateson and De Vries as the
true and only expression of permanent germinal modification. But, grant-
ing the truly germinal origin of mutations, it does not follow that they are
the only product of germinal modification.

A discontinuous variation, as the name suggests, is unconnected by inter-
mediate conditions with the usual (modal) condition of the species. It
represents a change, more or less abrupt, from the modal condition of the
species, and is strongly inherited, a fact which indicates clearly its exclu-
sively germinal origin.

In the category of discontinuous variations belong abrupt changes in
pigmentation and hairiness among both animals and plants, changes in the
number of digits or of the number of phalanges in a digit among vertebrates,
in the presence or absence of horns among animals and spines among plants,
and other similar conditions.

IN RATS AND GUINEA-PIGS. 3

Such changes are not the result of selection; they often appear, as it
seems, spontaneously, and they are permanent in the race, if isolated.

De Vries maintains that all species-forming variations are of this sort;
that selection is unable to form new species, because it can neither call into
existence mutations nor permanently modify a race by cumulation of
abmodal fluctuations. Darwin, on the other hand, and the great majority
of his followers, while admitting that races are occasionally produced by
discontinuous or "sport" variation, ascribe evolutionary progress chiefly to
the cumulation through long periods of time of slight individual differences,
such as De Vries calls fluctuations. The issue between the two views is
sharp and clear. According to De Vries, if we rightly understand him,
selection is not a factor in the production of new species, but only in their
perpetuation, since it determines merely what species shall survive; accord-
ing to the Darwinian view, new species arise through the direct agency
of selection, which leads to the cumulation of fluctuating variations of a
particular sort.

De Vries and the Darwinians differ not only as to the part which selection
plays in evolution, but also as to the nature of the material upon which
selection acts. According to De Vries, species are not modified by selection;
mutations are new species and selection determines only what mutations
shall survive, fluctuations having no evolutionary significance. On the
Darwinian view, all species, whether arising by mutation or not, are subject
to modification by selection.

A great deal can be said in favor of each of these contrasted views, but dis-
cussion is at present less needed than experimental tests of the views outlined.
To De Vries we owe much for showing that such tests are possible.

It was our purpose to make tests of this sort when we undertook the
experiments described in this paper. The questions to which principally
attention has been directed are these: (i) Can discontinuous variations be
modified by selection alone? (2) Can discontinuous variations be modified
by cross-breeding? A negative answer to these questions will support the
view of De Vries ; an affirmative answer will support the Darwinian view,
because it will show that through selection new conditions of organic stability
can be obtained; that is, new species may be produced.

The material used consisted of certain discontinuous variations in the
color-pattern of rats. The general result obtained is this: Various color-
patterns, like the several pigments found in the rodent coat, are mutually
alternative in heredity. Each group of individuals referred to the same
type of color-pattern forms a continuous series fluctuating in accordance
with the laws of chance about a common modal condition. The different
types in general do not overlap ; they form a discontinuous series. Now, these
types may be modified in two different ways : (a) By selection of abmodal

4 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

variates within the same continuous series, and (6) by cross-breeding between
different types. There is no evidence that one of these methods has effects
less permanent than the other. So far, then, as these experiments go, they
support the Darwinian view rather than that of De Vries.

VARIATIONS OF RATS IN COAT-COLOR AND COAT-PATTERN.

Variations in the pigmented coat of rodents are of two principal sorts:
(i) Variations in the character of the pigments found in the coat, and (2)
variations in the distribution of those pigments. The character of the
pigmentation in the wild rodent is nearly always complex. Two or three
pigments are associated together in the same hair, but they differ in their
regional distribution, so that a grayish or brown "ticked" coat results, incon-
spicuous against many backgrounds. The coat of the house-mouse (Bate-
son, 103) and that of the wild guinea-pig (Castle, 105) contain three optically
different pigments — yellow, brown, and black. These all coexist in the
same hair. In certain fancy varieties of these rodents, a single pigment is
present without the others, or the distribution of the pigments is such that
only one sort is conspicuous. Animals pigmented thus are known as black,
chocolate, and yellow (or red) varieties. If all three pigments are absent
from the coat and likewise from the retina of the eye, a condition known as
total albinism obtains.

In rats, rabbits, and certain other rodents, black and yellow self- varieties
are well known, but no pure chocolate animal has been observed.

Total albinism and the several "self" conditions of pigmentation are all
mutually alternative in inheritance.

Variations in pigment distribution on the body result commonly either
from entire absence of pigment from certain regions of the body, in which
case the coat has white markings, or from the occurrence of different pig-
ments singly in different body-regions, in which case the body bears spots of
different colors. Both sorts of variation may occur simultaneously, in which
case the body is spotted with pigments of different sorts and with white.

The color-varieties of rats are fewer and simpler than those of mice, rab-
bits, and guinea-pigs. Aside from albinos, there are only two "self" (i. e.,
uniformly colored) varieties, namely, gray (or brown, the color of the wild
Mus decumanus) and black. Gray is a Mendelian dominant in relation to
black.

As regards coat-pattern, there occur two conditions of partial albinism,
which differ from each other only in degree, but which may be obtained
each probably in a pure (homozygous) condition. These two patterns may be
called "Irish" and "hooded." Bach occurs either with gray or with black
pigmentation. The "Irish" of fanciers, as described by Doncaster (:o6),

PLATE 1

VARIATION IN EXTENT OF PIGMENTATION AMONG CROSS-BRED
HOODED RATS.

The usual condition in the original stock is shown in Fig. 4, reduced pigmen-
tation in Figs. I and 3, and increased pigmentation in Fig. 2.

THE MERIOEN GRAVURE CO.

IN RATS AND GUINEA-PIGS. 5

applies only to black-pigmented individuals, but for lack of a better descrip-
tive term, we shall apply it to all animals having the color-pattern of
"Irish" rats, whether gray or black pigmented. A rat of the. Irish pattern
has pigmented sides and dorsal surface, but bears more or less white fur
upon its belly, varying in extent from a few white hairs midway between
the front legs to a wholly white ventral surface.

Doncaster (:o6) has distinguished two types of Irish rats, in one of which
the white is more extensive than in the other. He has observed that the
rats with larger white areas are regularly heterozygous, producing hooded
as well as Irish offspring when mated inter se. Our experiments in the main
confirm this idea. It is not possible to determine with certainty, from the
size of the ventral patch alone, whether a particular Irish rat does or does
not contain the hooded pattern in a recessive condition, but in a lot of Irish
rats of similar ancestry those with the larger white patch oftener transmit
the hooded condition, while those which do not transmit the hooded condi-
tion oftener have a small white patch.

In hooded rats (pi. i, figs. 1-4) the white areas are more extensive than
in Irish ones ; pigment occurs only on the head, shoulders, and forelegs (con-
stituting the "hood" of fanciers), and as a median dorsal stripe extending
back on to the tail, the stripe being sometimes continuous, sometimes inter-
rupted, and of varying width. According to the nature of the pigment
which they bear, hooded rats may be distinguished as gray hooded or as
black hooded, precisely as animals bearing the Irish pattern are designated
either gray Irish or black Irish.

When crosses are made between rats differing in color-pattern, the more
extensively pigmented pattern tends to dominate in the offspring, a fact
recognized by Crampe ('77-84, '85), Bateson (:o3) and Doncaster (:o6).
The dominance, however, is not complete, so that the result might be de-
scribed as "goneoclinic," i. e., intermediate between the parental forms but
approximating one much more nearly than the other, in this case always
the more heavily pigmented one. Thus, a cross between a wild gray male
rat and a black hooded female, known to be homozygous, produced a litter
of seven young, all gray, but with a small patch of white on the chest, vary-
ing in extent from merely a few white hairs to an area of 4 to 5 sq. cm. The
same wild male, mated with an albino female, produced a litter of young
similar in character, all with some white below. Bateson (103, p. 78, foot-
note) mentions a similar result obtained by Miss Douglas.

Again, a cross between an Irish and a hooded individual produces Irish
offspring. An example will be found on page n in the matings of black
Irish $141 with black-hooded males, producing nineteen offspring, all Irish.
Most matings of Irish with hooded rats have in our experiments produced
offspring of both sorts, not because of reversal in the nature of the domi-

6 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

nance, but because the majority of the Irish rats used happen to have been
heterozygous, bearing the hooded pattern as a recessive character; but, as
we have seen, this is not a necessary relationship. Hooded rats, however,
bear no other color-pattern, so that any pair of hooded rats of whatever
pedigree will produce only hooded pigmented offspring.

To sum up in a brief statement the principal facts presented concerning
the coat-patterns of rats, the following series of conditions may be recog-
nized, each with pigmentation less extensive than the preceding: (i) Self,
whole body pigmented; (2) Irish, whole body pigmented except more or
less of the ventral surface ; (3) Hooded, only the head, shoulders, and usually
a median dorsal stripe pigmented; (4) Albino, no pigmentation.

Each condition behaves as a dominant toward those which follow it in
the series, and as a recessive toward those which precede it. Bateson (103,
p. 81) refers to Crampe's failure to obtain the self or the albino conditions
by selection from the parti -colored, and adds: "The types are in fact definite,
and can not be built up by cumulative selection." The statement applies
strictly only to the two extremes of the series, viz, self and albino, but by
implication to the others also. Our experiments, however, indicate that it
is possible to modify by selection and cross-breeding both the Irish and the
hooded conditions, leading to the production of intermediate conditions.
We suspect that the same may be true of the self condition also. It has
been pointed out elsewhere (Castle and Forbes, :o6; Castle, :o6), that the
theoretical absolute gametic purity, in Mendelian inheritance, probably does
not exist in any case. Heterozygosis leads inevitably to modification in
character of the conjugating gametes, not simply as regards the entire
assemblage of characters, but as regards each character considered sepa-
rately. The degree of modification is probably indicated roughly by the
imperfection of dominance in the heterozygote. Thus, when a perfect
blend is obtained, as in the inheritance of ear-length in rabbits, the gametes
transmit that intermediate character. But when dominance is very com-
plete, as in a cross between a self and an albino mammal, the segregation of
self and albino conditions is very complete among the gametes formed by
the cross-breeds. Yet when spotted individuals result from cross-breeding
between self and albinos, these spotted individuals form gametes which
transmit that same mosaic character. Now, the cross between self and
albino rats gives rats of Irish pattern, but quite variable. But our experi-
ments indicate that from variable Irish rats one can by selection obtain
rats transmitting no pattern but Irish. The theoretical importance of this
is obvious. Cross-breeding and selection combined are means by which
we may not only modify existing Mendelian characters, but may even
create new ones. They are, then, factors of prime importance in evolution,
even in the case of characters which vary discontinuously.

IN RATS AND GUINEA-PIGS.

MENDELIAN INHERITANCE OF COAT-COLORS AND COAT-PATTERNS

IN RATS.

While pigment character and color-pattern are both inherited in Mendelian
fashion, the two are entirely independent of each other. They are separate
and uncorrelated unit characters. Accordingly we find that each type of
pigmented coat-pattern, namely, self, Irish, and hooded, may occur either
with gray or with black pigmentation, the frequencies with which they occur
in the respective combinations being governed by the laws of chance and of
Mendelian dominance. This will appear in the detailed discussion of the
experiments.

Not only may each coat-pattern occur either in a gray or in a black
pigmented individual, but it may occur also in an unpigmented animal. Para-
doxical as this statement may seem, it is capable of abundant proof. The
coat-pattern of course is not visible in an unpigmented (albino) rat, but its
presence there as a potentiality can be demonstrated as certainly as the
occurrence of a recessive character in a heterozygous dominant individual.
Nothing but the presence of pigment is necessary to make the color-pattern
manifest. This can be supplied by a mating with a pigmented animal.

Specific pigment potentialities (gray, black, or both) are likewise present
in albinos. Consequently we must recognize that albinos transmit inactive
both pigments and color-patterns; these, however, are unseen and can not
be made visible until some lacking substance borne by all pigmented indi-
viduals is supplied. Characters transmitted in this inactive state have been
termed by one of us (Castle, : 05) latent, and that terminology will be followed
in the present paper.

ALBINISM RECESSIVE IN RELATION TO AU, TYPES OF PIGMENTATION.

That total albinism behaves as a recessive Mendelian character has been
recognized independently by a number of investigators, among the earliest
being Correns (: 01) and Cu^not (: 02). The fact has been abundantly veri-
fied in the case of mice (see Castle and Allen, 103), rats (Crampe, '77-84;
Bateson, 103; Doncaster, :o6), rabbits (Woods, 103; Hurst 105; and Castle
105), and guinea-pigs (Castle, 105). The experiments described in this
paper corroborate those of Crampe and of Doncaster with reference to rats.
The proportions of albinos and of pigmented individuals in mixed litters
are close to the Mendelian expectations, indicating neither selective union
of gametes nor lessened fertility of certain sorts of unions. Pigmented rats,
in which albinism was recessive, when mated inter se, have produced 129
albinos to 384 pigmented young, the numbers expected being 128 and 385,
respectively. Albinos mated with pigmented individuals, in which albinism

8 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

was recessive, have produced 201 albinos to 244 pigmented young, or 45.1
per cent albinos, where 50 per cent are expected. The deviation of nearly
5 per cent from expectation in the latter case is probably a matter of chance
and would grow less with more extensive observations, for in studying the
inheritance of total albinism some observers record an excess of albinos
(Castle, 105), others an excess of pigmented individuals (Allen, 104), while
others observe very close agreement with expectation (Cu^not, Darbishire).
This is as we should expect on the theory of probabilities (see Allen, 104, p. 82).

LATENT PIGMENT CHARACTERS AND COAT-PATTERNS.

Latent transmission of pigment characters through albinos is a matter
requiring fuller consideration. Assuming for the time being its correctness,
and knowing the well-established facts, (i) that gray pigmentation in rats
is dominant over black (Crampe, Bateson, Doncaster) and (2) that every
sort of pigmentation is dominant over albinism, we reach the following
conclusion. On Mendelian principles we may expect partial albinos to fall
into eighteen actually different classes, though visibly the classes number
only four, namely, gray Irish, black Irish, gray hooded, and black hooded.
These eighteen classes are enumerated in the four central columns of table i ,
page 34. Of gray Irish individuals, there should be eight actually different
classes as regards gametic output; of black Irish and gray hooded indi-
viduals, there should be four classes each; and of black hooded, two classes.

Albinos, though indistinguishable in appearance, should fall into nine
different classes as regards the latent pigment characters and color-patterns
which they transmit. The nature of these nine classes is indicated in the
last column of table i. The numerals prefixed to the class designations in
table i indicate the frequencies in which the various classes may be expected
to occur as a result of a particular sort of mating, namely, mating inter se
gray Irish individuals which bear recessive the three characters, total albin-
ism, black pigmentation, and hooded pattern, individuals designated
GI(W.BH). The table is based on the assumption that gray Irish indi-
viduals of the sort indicated are Mendelian triple heterozygotes, the three
independent pairs of allelomorphic characters being pigmentation versus
albinism, gray versus black, and Irish versus hooded. In support of the
assumption mentioned, it may be said that by proper breeding tests a majority
of the hypothetical classes have been shown to exist. Thus, of the eighteen
hypothetical pigmented classes enumerated in table i, all except four have
been shown to occur, those four being GI(W), GI(BH)* BI* and BI(W)*.
Further, five out of the nine supposed classes of albinos have been proved

* Demonstrated to exist since the foregoing was written.

IN RATS AND GUINEA-PIGS. 9

to occur, those the existence of which has not yet been demonstrated being
W[BI], W[GI]t W[G-IH], and W[GB-I]. It is probable that more extensive
tests would demonstrate the existence of all the missing classes, as no special
search has been made for them, the demonstrations obtained being for the
most part incidental to other investigations.

The fundamental assumption on which table i rests finds further justifi-
cation in the numerical proportions in which the various classes are observed
to occur, as will appear from an examination of tables 2 and 3 (pp. 35, 36).
In these tables are shown the theoretical Mendelian results and the actual
results of twenty- three different sorts of matings involving the pigments
and color- patterns which have been discussed in the previous pages. Table
2 includes only matings between pigmented animals; table 3, matings be-
tween a pigmented animal and an albino. Throughout both tables, it will
be seen, the expected Mendelian results agree quite closely with those actually
observed. Two animals possessing the same recessive character have in
no case produced offspring bearing the corresponding dominant character.
Further, the proportions of dominants and of recessives observed in mixed
lots of young agree in general quite closely with those expected. The most
striking deviation is an excess of black hooded young (BH) which is seen in
the totals for both table 2 and table 3. This, however, is not large enough
and does not occur with enough uniformity to warrant one in regarding it
as other than accidental. The totals for tables 2 and 3 combined are:

GI GH BI BH W Total.

Observed 69 132 130 186 168 685

Expected 56 137.5 136 158 187

In order of size the groups are as expected, except the last two, in case of
which the order is reversed.

In addition to the experiments recorded in tables 2 and 3, the following
observations may be mentioned as corroborative evidence of the Mendelian
behavior of pigment characters, color-patterns, and albinism in rats.

Black hooded individuals bearing white recessive, mated inter se, give
two classes of offspring, albinos and black hooded. These albinos have been
shown to bear without exception the black hooded pattern in a latent con-
dition. Of the black hooded offspring, there should be two classes — pure
hooded (free from albinism) and hooded individuals bearing albinism re-
cessive, the latter being twice as numerous as the former. Out of 22 off-
spring tested, 14 had albinism recessive and 8 were pure hooded individuals.

Matings of pure black hooded individuals with black hooded ones bearing
white recessive gave all black hooded offspring, 109 in number. One-half
of these should be pure hooded, and of the limited number tested this was
found to be true.

10

INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

The evidence contained in tables 2 and 3 (pp. 35, 36), is presented in the form
of summaries only. To give an idea of the sources from which it was obtained,
a genealogy may be considered in detail (fig. i).

Wild. oray

Wild white

J12,W[BB]

4 Ai,Gl(-

£ 34, G-lfw-BIf)

£ 26, Gl(w-Bff)

£«*

GI(W-B)

ventral white patch, below the average size

o 142
f 143
£144
J 145
<f 146
4 147

GI(H)
GI(W-BM)
G
BI(W-H)
GHfB)
BH

large "white "ventraX patch
gray self
large "white ventral patch

FIG. l.

A wild gray female and a wild white male were trapped together. Nothing
is known of their ancestry, but they were probably derived from escaped
tame albinos which had mated with wild gray rats. This pair of captured
rats produced gray Irish young, two of which (Ai and ,42), when mated with
ordinary albinos, produced gray Irish young which bore, as recessive char-
acters, total albinism, black pigmentation, and the hooded pattern. Two
of these triple heterozygotes ($34 and c?i6) were now mated together
and produced a litter of seven young (141-147, fig. i), which were tested
as to their gametic condition, with the results indicated in fig. i. Five of
the 7 were gray pigmented and 2 black, but no two were alike in gametic
character. Males 146 and 147 were mated with black hooded females and
with albino females of black hooded parentage, producing nothing but hooded
young. This result showed that they bore no coat-pattern except the hooded
one, and were free from recessive albinism.

The young of 6^147 were all black pigmented, showing that animal to
be entirely homozygous, but some of the young of 6^146 were gray pig-
mented, some black pigmented, showing that he bore black pigmentation
as a recessive character, being otherwise homozygous.

IN RATS AND GUINEA-PIGS. II

The gray Irish female (141) has produced, by hooded males, young of the
Irish pattern only, 19 in number, some of them being gray, and others black
pigmented; she has also produced albino young by hooded males bearing
recessive albinism. She accordingly bears black pigmentation and total
albinism as recessive characters, but is otherwise homozygous, for if she bore
the hooded pattern, half her pigmented young should have been hooded.
By similar tests, the gametic formulae of individuals 142 to 145 were estab-
lished, as given in fig. i.

Experiments such as these show the complete mutual independence of
pigmentation, color-pattern, and total albinism, as has been repeatedly
stated. They indicate, further, that partial albino rats (hooded and Irish)
regularly form gametes which bear the partial albino condition, either in
part or in all of their gametes; and that the extent of the white areas on
partial albinos varies in continuous fashion, a self condition being some-
times obtained as an extreme variation of Irish. This last point is impor-
tant as showing that no complete discontinuity exists between self and
Irish patterns in rats, since the gap ordinarily existing between them can
be bridged with a complete series of intermediate conditions. The same is
true of the Irish and hooded conditions. Though ordinarily discontinuous
and behaving as alternative Mendelian characters in inheritance, we can by
cross-breeding and selection bridge the gap between them.

MODIFICATION OF HOODED PATTERN BY CROSSING WITH IRISH.

Hooded and Irish individuals differ, as already stated, simply in the
extent of their pigmentation. In hooded individuals, the pigmentation is
restricted to the hood (which covers the head and shoulders) and to the
dorsal stripe; in Irish individuals, the entire coat is pigmented except a
ventral patch of variable size. Extension of the dorsal stripe of the hooded
pattern, so as to cover the entire dorsal surface and the sides of the body,
would yield the Irish pattern. It was our purpose in one set of experiments
to see whether such modification could be brought about, and to see also
how far an opposite modification of the hooded pattern (further reduction
of the pigmented areas) could be effected. Before considering those experi-
ments in detail, we may inquire what effect upon the hooded pattern a cross
with the Irish pattern will have.

The hooded stock used in this investigation consisted of individuals which
bore albinism recessive, but had no Irish ancestors, so far a? known. The
extent of the dorsal stripe varied considerably, so that it was found desir-
able to measure it as accurately as possible in each individual. This was
done, first, by estimating the width of the stripe in relation to the total
width of the back; next was estimated the extent to which the stripe was

12 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

interrupted, that is, the proportion of the total stripe length which was
pigmented. The product of these two ratios, relative width and relative
length of the back-stripe, when multiplied by 100, gives an approximation to
the per cent of dorsal surface (posterior to the hood) which was pigmented.
Such a valuation, or grade, was obtained for each hooded individual.

The grades for 183 individuals (lot A, fig. 2), belonging to three successive
generations, yield the variation curve shown in fig. 3, A . In obtaining this
curve grades from o to 9 inclusive were grouped in a single class, the mean
value of which is 4.5; grades from 10 to 19 inclusive, in a class the mean
value of which is 14.5, and so on. The ordinates in fig. 3, A, show the fre-
quencies of these classes. The grades fall into four classes, the one with
greatest frequency (modal class) being the one with the lowest mean (4.5),

the frequencies of the other classes
decreasing in order upward. The
average grade for the entire 183 in-
1 dividuals is 13.3 (see vertical line

r-J in fig. 3, A).

The hooded stock was now crossed
with an Irish stock, and the Irish
cross-breeds thus produced were in
j turn crossed back with the parental

t hooded stock. From this last cross

' , | ' ' were obtained both Irish and hooded

tj.-~-(uvi>jj'/-r4.j. H (JLtot C ) I ff . , , ,. f

i offspring, in proportions sensibly

H(LotE) equal. These hooded offspring form

a group (lot B, fig. 2) derived through
H (Lot F) one parent directly from the original

FIG. 2.- Ancestry of lots A to F of hooded rats. hooded Stock (lot A), but through
Read downward. the other parent from the Ifish cross.

breeds. One of the two gametes, accordingly, entering into each zygote had
a chance to be modified by association with the Irish character. Lot B
includes 126 individuals, with back-stripes on the average nearly twice as
extensive as those of lot A, the average grades of the two lots being 13.3
for lot A, and 21.2 for lot B. The form of the variation curve for lot B is
shown in fig. 3, B. The modal class is higher than in lot A, and the range
of variation is extended upward so as to include seven classes, the highest
one having a mean grade of 64.5. Accordingly, we conclude that a cross
with the Irish stock raises considerably the average size of the dorsal stripe
in hooded rats, as well as the range of variation upward in size of stripe.
In the same litters with the hooded rats forming lot B, were born rats of
two other sorts, Irish and albino (fig. 2) . The former bred inter se produced

IN RATS AND GUINEA-PIGS.

Individuals

KQ

1

\E

\

4-0

u

A

35

/v

/'

\ \
\
\

\

3O

7

\
\

\

\

a«T

/

y

\

?O

\

\

\

\
\
\

\

in

\

v

\

X

i

5"

*-£-

•--'

— -

"r

^^ V

V

\
\

\

^

i
i

sl

-*%

\(

•— -

— ,

2

>v,

-_

Grade 4.5

O(

\

fie

\

/sn

A
\

\

X

N\

ou

/'

\
\

\

4-f)

\

\

\

\\

J5

B,

X'

T"V" '4

\

\

F'

1

\\

5

kxv

1C

>j\

,^<

^>s

\

X

V

/"

XN

1

\

x-«

\

\

1 1

\

"•""""

w^'^

\

^ \

. ..

34.5 4-4.5 54.5 6*5 74.5

FIG. 3.— Variation in extent of the back-stripe in lots A to F of hooded rats, showing the effects
breeding with rats of Irish pattern. The positions of the means are shown by vertical lines

of cross-

14 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

the hooded rats forming lot C (fig. 2) ; the albinos, when back-crossed with
the original hooded stock, produced hooded offspring, which are included
in lot D (fig. 2), as are also the hooded descendants of lot B bred inter se.

Lot D, accordingly, is a group of individuals of ancestry more mixed than
lot B, but on the whole similar. It shows a similar variability (fig. 3, D),
but is of a higher average grade and shows a tendency to form a second
mode, a group of wide-striped individuals, with a modal grade of 54.5, i-e.,
individuals in which the stripe is about half as wide as the entire dorsal
surface (compare pi. i, fig. 2).

Lot C (fig. 2 and fig. 3, C) contained only 36 individuals. Its average
grade is almost identical with that of lot D, 23.5 as compared with 23.9.
Its range also is similar; but the curve itself is flatter, yet with the same
tendency to become subdivided into two groups, one with a wider stripe,
the other with a stripe like that of the original hooded stock (lot .4). Two
other lots of hooded rats (E and F, figs. 2 and 3) belonging to this series
have also been studied. Lot E consisted of the hooded young of certain
individuals of lot D, those having the lesser amount of Irish ancestry, while
lot F was derived from certain individuals of lot E bred inter se. The parents
of lot E had an average grade of 21.8, similar to that of lot B (21.2), or to
the lower modal group of lot D; the average grade of lot E itself was very
similar, namely 21.1. The parents of lot F had an average grade of 17.8;
lot F itself had an average grade of 12.9. The variation curves for lots E
and F are unimodal, like those for the original lot A, giving no evidence of
a tendency to form a wide-striped variety.

MODIFICATION OF HOODED PATTERN BY SELECTION.

The close agreement between the grades of the parents and children in
lot F, and measurably also in lot F, suggests that one can at will either in-
crease or decrease the width of the stripe in a stock of hooded rats. To
test this matter more fully, selection has been made both for reduced and for
increased size of stripe.

SELECTION FOR REDUCED STRIPE.

Eleven individuals of lot A, having narrow or interrupted dorsal stripes
(average grade n), formed the beginning of a selection experiment for
reduced pigmentation (compare pi. i, fig. 4). They produced 83 young,
lot G,* figure 4, with an average grade of 9.6, and a form of variation
curve suggesting the presence of two groups of individuals, each with a dorsal

*On account of the smaller range of variation in this series of experiments, the size of
the classes was made only half as great as in the case of lots A-F, fig. 3. The mean values
of the classes in lots G-I are 2, 7, 12, 17, etc.

IN RATS AND GUINEA-PIGS-

\

70

60

50

\

4-0

stripe of different extent, one group being of moderate size (12), the
other extremely reduced (2). Twenty-eight individuals whose grades fell
in the lower part of curve G (fig. 4), and averaged 7.8, were the parents of
the next generation, group 77. Individuals
This generation consisted of
6 1 individuals having an aver-
age grade of 5.6. The varia-
tion curve for this group (77,
fig. 4), shows a nearly com-
plete dropping out of individ-
uals with the stripe of moder-
ate size, such as occurred in
group G. All the classes
above 7 are small and the
upward range stops at 22. A
few individuals of lot 77 had
no back-stripe at all (com-
pare pi. i, fig. 3). These with
some individuals having a
much reduced stripe were se-
lected as parents for the next
generation, lot 7. The aver-
age grade of the parents was
in this case 1.7; that of their
34 young (lot 7, fig. 4) was 4.
Twenty-four of the 34 young
fell in the lowest class, mean
grade 2, while none were
above class 12.

The effect of selection in
this series of experiments is
clear from a comparison of
the four curves, A, G, 77,
and 7, fig. 4. Selection has
steadily lowered the average
amount of pigmentation in the^r<?c/e

race by reducing the Upward I?IG- 4-~Effect of selection for back-stripe of reduced size in
£ ^ • ±- hooded rats, lots A,G,H, and /.

range of the variation curve,

thus eliminating the more abmodal variations, reducing the skewness of the
curve, and diminishing the distance between mean and mode. One of the
noteworthy features of the case is the absence of what may properly be

17 32 87 33 37 +2

i6

INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

called regression. The filial average does not, with any uniformity, lag
behind the parental average, in the process of displacement downward in the
variation figure. This fact, together with the decreasing skewness of the
variation curve, indicates that the effects of selection in reducing the extent

Individuals

70

6O

\

V

\

30

2O

IO

GraCf€4.5 I4.S 24*5 34.5 44.5 5-4.5 64.5 74.5 64.5

FIG. 5— Effect of selection for back-stripe of increased size in hooded rats, lots A, J, K, L, and M.

of the pigmentation will be permanent, that is, that a stable, narrow-
striped variety of hooded rats can be established by selection, and that
this variety will breed true-

IN RATS ANt> GUINEA-PIG§. t^

SELECTION FOR STRIPS OF INCREASED SIZE.

Selection in an opposite direction, for a wider and more continuous stripe,
leads to similar conclusions. In this series of experiments, the wide-striped
parents used were obtained as a result of crosses between hooded and Irish
individuals. Such crosses tended to widen the stripe of the hooded offspring,
the widening effect being permanent, as the sequel shows.

Hooded individuals, either parents of wide-striped young, or such as
themselves had wide stripes, were selected from lots B and C, and mated
together. The average grade of the parents was 12. Their young (lot /,
fig. 5), in in number, had a higher average grade than the parents, namely,
15.8. The curve was similar in form to the curve (4) produced by the
original hooded stock.

A second group of wide-striped individuals, of higher grade than that
just described, though of similar ancestry, produced a group of 70 young
(lot K, fig. 5), like themselves in width of stripe. The average grade of
the parents was 23, that of the young (lot K) was 22.2. The variation
curve (K, fig. 5) is a very flat one, similar to curve B, fig. 3.

A third group of wide-striped parents, of still higher grade, was taken
chiefly from lot D. Their average grade was 40.7. They produced 64
young (lot L, fig. 5), having an average grade of 36.2. The form of variation
curve given by lot L is similar to that given by lot D, flat and bimodal.
The upper mode, however, is relatively taller, indicating that the upper
(or wide-striped) group is increasing at the expense of the lower, as a result
of the selection, and that the original two-peaked condition of curve D was
due, not to the heterogeneity of the material included in lot D, but to the
fact that part of the gametes formed by the cross-bred individuals trans-
mitted a modified (wide-striped) condition. This wide-striped condition
was now (in lot L) in process of segregation through the action of selection,
and in a fair way to form a stable wide-striped race. Indeed, in the best
wide-striped pairs of lot L this seems already to have been accomplished;
for certain extremely wide-striped individuals (partly of lot L and partly
of a later generation), which had an average grade of 50, produced the 19
young included in lot M, fig. 5, which had an average grade of 53. The
mode for this curve is close to the mean, being 54.5, and the curve itself
is nearly symmetrical, indicating approach to a condition of stability.

The theoretical importance of the results of these selection experiments
is evident. Characters can be permanently modified by selection, contrary
to the view of De Vries that such modification is impossible.

The objection may be raised that the permanency of the modification has
not been fully established, for later generations might show regression to
a narrower-striped condition. While admitting that this objection has

1 8 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

some force, the fact may be pointed out that regression does not occur in
the experiments made, and if regression were to occur at all, we should expect
it to occur and to be greatest in the early stages of a selection experiment.
The average grade of the offspring is sometimes greater, sometimes less, than
that of their parents, as we should expect if these deviations are purely the
result of chance. Table 4 (p. 36) shows the nature and amount of the
deviation following each selection made. No great importance attaches to
the exact amounts of deviation indicated in this table. For in computing
the parental averages all parents were weighted alike, but a better method
of procedure would have been to weight each parent in proportion to the
number of young which it produced. The general nature of the result
obtained is, however, not vitiated by the procedure followed; it serves to
show the absence of any uniform regression of either a plus or a minus nature.

INDIVIDUAL SPOTS OF GUINEA-PIGS NOT UNIT CHARACTERS.

In an earlier paper (Castle, : 05) the fact was pointed out by one of
us that in partial-albino guinea-pigs, as in other mammals, the pigmented
areas are not distributed without order over the body, but occur in regions
fairly definite. These regions in the guinea-pig consist of live paired "pig-
ment centers," dorso-lateral in position. To define their location more
accurately, they have been called the eye, ear, shoulder, side, and rump areas.

The hypothesis was advanced (Castle, : 05, p. 45) that these several areas
might represent separately heritable unit characters, although experience
had clearly shown that, if such were the case, they were not inherited in
ordinary Mendelian fashion. To test this hypothesis more fully, we have
attempted to establish races of guinea-pigs which should possess certain of
the pigmented areas, but should lack others. For this purpose experiments
have been made with three different patterns of partial albinism, arbitrarily
chosen from those which occur among ordinary spotted guinea-pigs. These
are the Dutch-marked pattern (pi. 2, fig. 3), the head-spot pattern (pi. 2, fig. 2)
and the nose-spot pattern (pi. 2, fig. i).

COLOR-PATTERNS SELECTED.

In the Dutch -marked pattern, all the typical spots are present except the
shoulder spots. The absence of shoulder spots produces a white belt or
girdle round the body. Anterior to the girdle, eye and ear spots are present
and usually confluent on the same side of the head, but not across the median
plane. Posterior to the girdle, the large side spots and the small rump spots
are present and confluent over the back as well as lengthwise of the body,
forming a large blanket of pigment covering the whole posterior part of the
body, but not extending out upon the feet. This is the ideal "Dutch-
marked" type of the fancier. Occasionally it is fully realized, but more
often minor departures from the ideal occur, such as asymmetry of the

PLATE 2

COLOR-PATTERNS SELECTED IN EXPERIMENTS WITH GUINEA-PICS.

Fig. I. Nose-spot pattern (Series N).
Fig. 2. Head-spot pattern (Series H).
Fig. 3. Dutch-marked pattern (Series D).

The rough (resetted) condition of the coat shown in Figs. I and 2 is a character
wholly independent of color-pattern.

IN RATS AND GUINEA-PIGS. ic)

pigmentation of the two sides of the body, failure of the eye and ear spots
to coalesce, or entire absence of one of the two, absence of one or both side
spots, or the occurrence of a white gap between side and rump spots. Our
attempts to fix this pattern by selection, continued through a series of gen-
erations, have not been successful.

Animals of the head-spot pattern (pi. 2, fig. 2) differ from Dutch-marked
chiefly in the entire absence of pigmentation posterior to the girdle, though
the pigmentation on the side of the head also is usually less extensive than
in Dutch-marked animals. Because of the practical difficulty of distin-
guishing eye spots from ear spots, when the two are confluent and similarly
pigmented, we have throughout this series of experiments treated the two
as forming a single pair, which we call head spots.

When the head pigmentation is most reduced, only the retina of the eye
is pigmented. Then the head-spot type passes over into the black-eyed
white type. In tabulating our results, however, we have treated the two
types as distinct. The head-spot pattern, like the Dutch-marked, we have
been unable to "fix."

The nose-spot pattern occurs less frequently than the two already de-
scribed. The entire body is white, except that part of the head which is
situated anterior to the eyes. This part of the body is usually unpigmented
in spotted guinea-pigs, which fact makes the occasional occurrence of the
nose-spot pattern more striking. Our attempts to fix this pattern have
succeeded scarcely better than those with the Dutch-marked and head-spot
patterns.

In addition to our attempts to fix particular color-patterns arbitrarily
chosen, we have made records of the pigment distribution on several hun-
dred spotted guinea-pigs of known ancestry, with a view to ascertaining
whether or not particular spots are inherited. The question for which an
answer was sought may be stated thus: Does an animal having, let us say, an
eye spot, produce young with eye spots any oftener than animals bearing a
like amount of pigmentation but without eye spots ? If this question receives
an affirmative answer, then we may conclude that eye spots, as such, are
inherited, and it will remain to ascertain the law governing their inheritance.

The evidence bearing on this question can best be presented in the form
of tables, showing what areas of the young are pigmented in comparison
with those of their parents. Such a table has been constructed for each of
the successive generations in which selection was exercised for the fixation
of a particular coat-pattern. By a comparison of the tables for the succes-
sive generations, we may learn whether selection was having any effect in
fixing the desired coat-pattern. For, if it was doing so, later generations
should show closer agreement than earlier ones between parents and children
as regards distribution of pigment on the body.

20 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

DUTCH-MARKED SERIES (SERIES p).

The data for inheritance of spots in the Dutch-marked series are con-
tained in tables 5 to 14. The individuals are arranged in a series of groups,
each group having a larger average amount of Dutch-marked ancestry than
the foregoing one. For convenience in description, we may use the symbol
D to designate any Dutch-marked animal, followed by a numeral to indicate
the amount in generations of its Dutch-marked ancestry. On this basis an
original Dutch-marked animal, one without Dutch-marked ancestors, would
be designated DQ; a Dutch-marked animal with one generation of Dutch-
marked ancestry, D\ ; one with two generations of Dutch-marked ancestry,
D2 ; and one with three generations of Dutch-marked ancestry, D3 . Simi-
larly a Dutch-marked animal, one parent of which was Dutch-marked, but
the other not, might be designated D\ ; the Dutch-marked young of D\
individuals might be called D\\ ; etc.

The young of our Dutch-marked animals have been divided into four
groups, the average amounts of Dutch-marked ancestry in these groups
being respectively ^, i, 2, and 3 generations. Groups of individuals were
at first created which came midway between the groups mentioned, but
these differed so little from the adjacent groups that we decided finally to
abolish them, dividing them equally between the groups to which they
were intermediate. Thus a group, DI%, was divided equally between the
groups DI and D2, and a group, D2\, between the groups D2 and D3.

For each individual observed, a record was made of what spots it possessed
and what spots its parents possessed. These records are shown in tables
5, 7, 9, and 1 1 for the four groups of individuals having the mean ancestral
values D$, DI, D2, and Z>3 respectively. Bach table contains separate records
for head, shoulder, side, and rump spots. The significance of the various
entries will be sufficiently clear after an examination of a single category
(head spots) in table 5. There are for the individuals examined three pos-
sible conditions as regards head spots. An individual may have two such
spots (one on either side of the head) ; it may have only one head spot
(right or left, as the case may be) ; or it may have no head spots. Records
are made for these three conditions separately in the horizontal columns
marked 2 spots, i spot, and o spots respectively. For the possible head-
spot conditions of the two parents considered jointly, five categories are
necessary, 4, 3, 2, i, and o. These are indicated in vertical columns in the
tables. An entry in the upper left-hand square of the table signifies an
individual having a pair of head spots, born of parents both of which were
similarly marked; an entry in the next square to the right means an indi-
vidual having likewise a pair of head spots, but born of parents only one of
which had a pair of head spots, the other having one head spot only. The

IN RATS AND GUINEA-PIGS. 21

table, accordingly, it will be observed, represents the summation of a large
amount of detailed information concerning the distribution of spots in off-
spring and parents, respectively.

For convenience in comparing the categories of individuals one with
another, the entries in the tables are expressed in per cents, as well as in
numbers of cases observed. Thus in table 5, left-hand vertical column,
entries are made concerning the head-spot conditions of 49 young born of
parents both of which had two head spots. Of these 49 young, 42, or 85.7
per cent, had two head spots, as each of the parents had; 6, or 12.3 per cent,
had only one head spot; while i individual, or 2 per cent, had no head spots.
The other columns of the table are to be interpreted in a similar way.

Glancing over the four divisions of table 5, which includes observations on
1 1 7 young, we notice that the offspring of parents both of which possess a
particular pair of spots (left-hand vertical column) possess that same pair
of spots in from 70 to 86 per cent of all cases. The agreement between
parents and offspring is greatest (85.7 per cent) in the case of head spots,
and least (70 per cent) in the case of side spots.

From the last vertical column of table 5 we learn that when both parents
lack shoulder or side spots, 50 to 58 per cent of their young lack the corre-
sponding pair of spots entirely.

These facts seem to indicate a rather strong degree of inheritance of par-
ticular spots upon the body of the guinea-pig; so we at first interpreted
them, but more extended observations limit this conclusion considerably.

The four divisions of table 5 are combined in table 6, which accordingly
shows, for the body as a whole, to what extent particular spots are inherited
in the group of 117 individuals with which tables 5 and 6 deal. Inheritance
of paired spots is seen to be, on the average, measured by 78.8 per cent (left-
hand upper square, table 6) ; absence of paired spots is seen to be measured
in inheritance by 55 percent (lower right-hand square, table 6). Further,
the fewer spots the parents have, the fewer their young have, as a glance at
the horizontal columns of table 6 will show. This is true not only of the
body as a whole (table 6), but also of its individual parts (table 5). Par-
ticular amounts of pigmentation are accordingly inherited quite strongly;
distribution of that pigment in certain of the typically pigmented areas,
rather than in others, is really not inherited, as we shall see, though such
is the impression first given by these tables.

For the second group of individuals of Dutch-marked ancestry (mean
ancestral condition, DI), the observations concerning spot inheritance are
given in tables 7 and 8; for the third group (D2), in tables 9 and 10; and for
the fourth group (#3), in tables n and 12. The numbers of individuals in
these several groups are 30, 178, and 136, respectively, making in the Dutch-
marked series as a whole 461 individuals observed.

22 INHERIT ANCE OF COAT-PIGMENTS AND COAT-PATTERNS

To obtain averages for the entire series, with which the values for the
several groups might be compared, tables 5, 7, 9, and 1 1 have been combined
to form table 13, which accordingly shows the average degree of inheritance,
in the entire series, of head spots, shoulder spots, side spots, and rump
spots, respectively. Similarly, tables 6, 8, io,and 12 have here been combined
in table 14, which accordingly shows in general to what extent paired spots
are inherited in the Dutch-marked series.

A comparison of the inheritance coefficients (per cents) in corresponding
places of tables 5, 7, 9, and n shows remarkably little change from one
generation to another. Paired head spots are inherited in the four groups
in 85.7, 82.1, 86. 5, and 89.3 per cent of the cases, respectively, or in the series
as a whole, in 87 per cent of all cases (table 13, upper left-hand square).
Similarly, paired rump spots are inherited in the four groups in 7 1 .4, 80,
80.9, and 83.1 per cent of the cases, respectively, or in the series as a whole
(table 13) in 80.7 per cent of all cases. In both head and rump spots the
inheritance coefficients increase slightly with the progress of selection.

Shoulder and side spots are so frequently lacking in the Dutch-marked
series that comparison of the per cents for presence of these pairs would
have little value; it is better to examine rather the per cents for absence
of these spots. Both shoulder spots are absent in the four groups in 58.5,
59, 41, and 42 per cent of the cases, respectively (a decreasing series) ; average
for the four groups, 46.3 per cent. Similarly, side spots are entirely wanting
in the several groups in 50, 34.7, 47.2, and 44.8 per cent, respectively (again,
on the whole, a decrease), the average for the four groups being 46 per cent.

It is scarcely necessary to examine in detail other parts of the several
tables. They show conditions for the most part intermediate between those
seen in these extreme instances.

Selection for presence of spots (head and rump) and for absence of spots
(shoulder) seems, in the same series of observations, to give contrary results.
These seemingly inconsistent facts are really not inconsistent. Spots
desired present and selected for have become slightly oftener present; spots
desired absent and selected against have likewise become slightly oftener
present. Selection for particular spots has had nothing to do with bringing
about the result observed. The pigmentation has become somewhat more
extensive in the race, without our having consciously aimed to make it so.
Spots occur somewhat more numerously in the later than in the earlier gen-
erations of the series (see table 26). The increase, however, takes place no
oftener in regions where increase is desired (head and rump) than where it is
not desired (shoulder) , or where its occurrence is a matter of indifference (side) .

This result shows the entire inefficiency of selection in guinea-pigs to fix
a coat-pattern which is dependent upon the distribution of pigment in par-
ticular spots upon the body.

IN RATS AND GUINEA-PIGS. 23

Head spots regularly give higher inheritance coefficients than any other
set of spots, a possible reason being that head spots really include two
distinct pairs of pigment areas, namely, eye spots and ear spots. Commonly,
in the Dutch-marked series, both of these are pigmented, in which case it
makes no difference whether the two are dealt with separately or jointly-
But sometimes an eye area may be pigmented, while the adjacent ear area
is unpigmented, or vice versa; in such cases separate enumerations of eye
and of ear spots would indicate a less amount of head pigmentation than
the joint reckoning followed, for purposes of convenience, in this paper.
Aside from this consideration, however, it is certain that the head region is
oftener pigmented than any other part of the body.

HEAD-SPOT SERIES (SERIES H).

The data concerning spot inheritance in the head-spot series are contained
in tables 15 to 24. The entries are made separately for the different regions
of the body in tables 15, 17, 19, and 21, for four different groups of indi-
viduals possessing different average amounts of head-spot ancestry. The
average amounts of head-spot ancestry in these several groups are a half-
generation, one generation, two generations, and three generations, respec-
tively, and may be expressed by the symbols //£, H\t H2t and H3) as explained
in connection with the Dutch-marked series. In the group H\ (table 15),
one of the parents of each individual recorded bore certain spots other than
head spots, but in the other three groups of Series// (tables 17, 19, and 21),
none of the parents bore spots other than head spots. Tables 15, 17, 19,
and 21 are combined in table 23, which shows the average degree of in-
heritance of particular spots in the series as a whole.

As in the case of Series D, the entries for the different regions of the body
have, in the case of each group of individuals, been combined to get an expres-
sion for spot inheritance in general, in the particular group of individuals
under consideration. These combination entries are found in tables 16, 18,
20, and 22, for the four groups considered separately, and combined into
one table for the whole head-spot series, in table 24.

In this series, rigid selection has been exercised for head spots, and against
all other spots. We will now consider what the effects of this two-fold
selection are. Paired head spots are present in the following per cents of
the young of parents possessing such spots, in the several groups: 79.8,
85.4, 77.1, and 82.1 per cent; average for the series (table 23),8r.i per cent.*
No change is observed uniformly in the direction either of increase or of

*In obtaining the average, each table is weighted in proportion to the number of individ-
uals recorded in it. These are, for the successive groups, 207, 193, 129, and 58; total, 587.

24 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

decrease of head spots, but they are on the whole less common in this series
than in Series D (81.1 and 87 per cent, respectively).

However, shoulder, side, and rump spots become steadily less common in
the series, as a glance at table 25 will show. It is doubtful, however, whether
this is to be regarded as an effect of selection for elimination of specific areas.
It indicates, rather, a decrease in the average amount of pigmentation pos-
sessed by animals of this series, unwittingly brought about by the attempt
to eliminate most of the spots typically pigmented (compare table 27). It
is significant that head spots, though desired present and chosen equally
in Series D and in Series H, occur more commonly in the former than in the
latter. Animals of Series D are more extensively pigmented than those of
Series H (compare tables 26 and 27) ; the additional pigmentation may occur
in any of the typically pigmented areas, regardless of the areas which happen
to have been pigmented in the selected ancestors.

Nevertheless, we find a more rapid falling off in the pigmentation of some
areas than in that of others. In Series H, shoulder spots fall off 16.9 per
cent; side spots, 7.3 per cent; and rump spots, 6.4 per cent; while head
spots, as already noted, show little change. Now, selection, it will be remem-
bered, was exercised for elimination of shoulder, side, and rump spots alike,
but for the retention of head spots. Accordingly we do not get changes
uniformly consistent with selection. Regardless of selection for specific
spots, we get, when the total amount of the pigmentation is decreasing, a
decrease most rapidly in shoulder spots, less rapidly in side spots, less rapidly
still in rump spots, while the head^spots are least affected. Conversely, in
a series in which the amount of pigmentation is increasing, the increase is
less marked in head than in rump spots, and less in side than in shoulder
spots. Thus, in Series D, head spots increase 3.6 per cent, rump spots,
11.7 per cent; the change in side spots (entire absence) is 6.2 per cent, of
shoulder spots, 16.5 per cent.

In further support of this same idea — that with increasing or decreasing
pigmentation changes occur more rapidly in some areas than in others,
irrespective of the particular spots borne by the ancestors — we may com-
pare the averages for Series D and H (tables 13 and 23). The differences
between the two series are, in frequencies of head spots, 5.9 per cent; of
side spots (absent), 20 per cent; of shoulder spots (absent), 25.4 per cent.
As regards rump spots, no comparable entries occur in either the first or
the last columns of tables 13 and 23, but an examination of the middle column
(2) in both tables indicates that the difference in rump spots in the two
series is about as great as in head spots.

These facts are in harmony with an observation made in advance of the
statistical investigation, that the most persistent of all the pigmented areas
are the head spots, and next in order of persistency are the rump spots.

IN RATS AND GUINEA-PIGS. 25

Our general conclusion from a study of these two series of guinea-pigs, includ-
ing 1,048 individuals, is that one can by selection either increase or decrease
the extent of the pigmented areas, but it is impossible by selection to fix this
pigmentation in a particular pattern, retaining pigment areas on certain parts
of the body and eliminating them from others. As the pigmentation changes
in extent, under the influence of selection, the various areas typically pigmented
are affected in the following order: Shoulder, side, rump, and head, the change
being greatest in the first-named and least in the last-named area, irrespective
of what particular spots were present in the selected ancestors.

STATISTICAL ANALYSIS OF THE DATA FOR SERIES D AND H.

The foregoing conclusions may, we think, fairly be drawn from an exami-
nation of the tables as they stand. For those, however, who place confidence
in the more precise methods of statistical analysis devised by Pearson and
others, it may be more satisfactory to treat the tables, which have been
constructed for the various groups of individuals, as correlation tables, and
derive from them the constants which measure the variability of parents
and children respectively in the several groups, and the degree of correlation
between the two. Such constants are given, in tables 26 and 27, for the
several groups of Series D and H, respectively.

From the left-hand columns of table 26 we learn that, in Series D, the
number of spots borne by an individual increases from group to group, in
the case of both parents and offspring; but the number of spots is, in every
case, 20 to 26 per cent greater in the offspring than in their parents. This
indicates a tendency for the offspring to become pigmented in regions which
lacked pigment in their selected parents.

Table 27 shows the existence of a similar, but still stronger, tendency in
Series H, the offspring bearing 50 to 105 per cent more spots than their
parents. This tendency may be considered regression, a tendency to return
to a condition of more widely distributed pigmentation, and to acquire
spots where the selected ancestors lacked them. It does not imply that the
young bear more pigment than their parents. The regression is stronger
the more reduced the pigmentation of the parents, as we might expect.
This is seen to be true both within Series H and in that series as a whole
compared with Series D.

The standard deviation from the average number of spots is for the
offspring a very constant quantity, being close to 3.6 in both Series D and
Series H. This indicates no change in the variability in number of spots
as a result of selection, or as a consequence of change in the number of pig-
mented areas. The parents show throughout Series H and in Group D% of
Series D a less standard deviation than their offspring. This fact, however,

26 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

has for our purposes little significance, since the parents are selected indi-
viduals, while their offspring are not. The same may be said of the coeffi-
cients of variability for the parents. For the offspring, the coefficient of
variability (ratio of standard deviation to mean, times 100) decreases slightly
in Series D, and increases in Series H. This must not be interpreted as
signifying a change in variability of an opposite nature in the two series.
It is due entirely to the changes in the mean (average number of spots),
these changes being opposite in nature in the two series. If the mean re-
mained constant throughout each series, the amount of variability indicated
by the coefficient of variability would likewise remain constant ; for we have
seen that the standard deviation is constant, irrespective of changes in the
mean and irrespective of the number of generations during which selection
has been in progress. Standard deviation is, therefore, a better measure
of variability than the coefficient of variability in these series.

The coefficient of correlation (r) has little real significance in tables 26
and 27. This constant is relatively small in the Dutch-marked series and
grows smaller as selection progresses, whereas in the head-spot series it is
relatively large and grows larger with the progress of selection. Selection
does not have an opposite effect in the two series upon the inheritance of
a coat-pattern. The whole effect is due to change in the amount of the
pigmentation, not to its distribution. When the amount of pigmented
surface is small, its distribution in certain spots (head and rump) is more
certain and the correlation rises (Series //). When, on the other hand, the
pigmented surface is large (Series D, table 26), its distribution is less certain
and the correlation is low; it becomes lower as the pigmentation increases,
in spite of progressive selection.

The statistical analysis confirms the conclusions previously drawn from
our observations. Selection is powerless to fix a particular coat-pattern
not dependent upon amount of pigmentation. It is as powerless to decrease
the variability in number of spots as to fix any pattern formed by them.

NOSE-SPOT SERIES (SERIES N).

It is only in the case of the nose-spot series (Series N) that we are able to
detect influence of selection in fixing a pigment-pattern among guinea-pigs.
Two individuals showing this rather striking variation in pattern were figured
by Castle (105, pi. 6); another is shown in pi. 2, fig. i, of this paper. The
starting-point of the series of selections for the fixation of this pattern was
the male 1989° shown in fig. 12, pi. 6 (Castle, 105). Besides the nose spot
shown in the figure, this animal bore a distinct right shoulder spot, not
visible in the figure. This animal produced a considerable number of young
with nose-spot markings similar to his own, as will be seen from an examina-
tion of table 28. As a result of matings with animals which did not bear

IN RATS AND GUINEA-PIGS. 27

nose spots, he produced four individuals with nose spots and fourteen with-
out nose spots, or 22.2 per cent of nose-spot (NS) young. When he was
mated with these nose-spot young (N S%) or with nose-spot individuals
produced in other experiments but not of nose-spot ancestry, and so desig-
nated N S0, he produced a much higher percentage of nose-spot young, viz,
53.6 per cent by AT So mothers, and 61.5 per cent by N S$ mothers. In
one case he was mated with his NSi daughter by an N S0 mother, from which
mating there resulted three nose-spot young and two otherwise marked, or 60
per cent N S. The number of young in this last experiment is too small
to be very significant quantitatively, but those in the other matings are
large enough. They show an increase in the per cent of NS young with an
increase in the amount of NS ancestry, and strongly suggested the possi-
bility of a still further increase by continued selection. Such increase,
however, has not up to this time been realized. Five sons or grandsons of
c? 1989° have been quite extensively tested and the same is true of one
individual ((^5595, NSi, table 29) descended from a brother of 1989°. Of
these six males, two have records, about equaling in production of nose-
spot young the record of c? 1989*, but the records of the remaining four
are inferior to his. The two males with the best records are ^5652 (table 29),
a son of cT 1989*, and cT 5669 (table 29), a grandson of c? 1989° and son of
c?5i5i (table 29). Neither of these males, however, has produced a large
number of young (the former 16, the latter 10), and so too much importance
must not be attached to the per cent results. Of all the six males, 5151
alone has produced more than fifty young. By NSQ and AT Si mothers alike,
about 40 per cent of his young bear nose -spots. His father's record by the
same group of mothers was about 15 per cent better (see table 28).

The collective results given by the six males enumerated in table 29 are
shown in table 30. They indicate that in the long run a higher percentage
of NS young is produced by NSi than by NS0 sires, but neither group
gives as high a percentage as the original nose-spot male, 1989° (table 28).

The nose-spot mothers employed in these experiments are much more
numerous than the males, but because the number of young produced by
any one of them is in no case greater than fifteen, the results are not given
for the mothers individually, but only in collective form (table 30, last part) .

The collective results for the females, like those for the males, show an
increase in the percentage of nose-spot young with increase in the amount
of nose-spot ancestry. In other words, the character is apparently inherited
feebly through both the male and the female lines, and is being gradually
fixed by selection. The process, however, is a slow one. After two gener-
ations of selection, the inheritance coefficient is no higher than in the original
male, 1989". Whether it can ultimately be made higher remains an open
question.

28 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

The objection may be made that a nose spot as such is not inherited,
any more than an eye spot or a rump spot (see page 25) ; that in a spotted
race a certain average amount of pigmentation is an inherited condition,
but the distribution of this pigment is wholly a matter of chance. To test
this point, comparison has been made with the young of certain sires in Series
H (table 31). In the case of the litters recorded in this table, neither parent
bore a nose spot, nor came from a nose-spot family. The extent of the pig-
mentation of the parents is measured roughly by the number of the typical
spot-areas which were pigmented (see columns 2, 3, 4, 7, and 10, table 31).
For comparison with the nose-spot series, see columns 3 and 6 of tables 28,
29, and 30. It will be observed that the extent of the pigmentation is similar
in the nose-spot and head-spot series, being in both cases close to an average
of three spots to an individual. The percentage of nose-spot young, how-
ever, is much lower in the eye-spot than in the nose-spot series, being 16.3
per cent in the former as compared with 40 to 50 per cent in the latter.
That the nose-spot marking does occur quite frequently among spotted
guinea-pigs not selected for that character is shown clearly by table 31.
That its occurrence, however, is more frequent when selection is made for
the character is evident from tables 28 to 30.

In table 31, as in tables 28 to 30, it will be observed that nose-spot young
have fewer other spots than do their brothers and sisters that lack nose
spots. The average difference is from half to two -thirds of a spot. This
means that a nose spot takes the place, to some extent, of pigmentation
elsewhere, and it is in part a matter of chance whether the pigmentation is
located on the nose or elsewhere. But chance is not the only element (if
we may so speak of chance) entering into the matter, for nose-spot parents
clearly produce more nose-spot young than do other parents transmitting
a like amount of pigmentation. How, then, are we to account for the fact
indicated by our observations that, while other patterns are unfixable, the
nose-spot pattern is in part at least fixable? At present we can not account
for it, but a consideration of familiar facts concerning mammalian develop-
ment may help us in shaping a hypothesis.

The production of hair and skin pigments in guinea-pigs is the exclusive
function of the ectoderm, as shown by Leo Loeb ('97) and confirmed by
Castle (-.05). In spotted guinea-pigs the limits of the pigment spots are
very precisely defined at birth and these limits, so far as we have been able
to observe, are never subsequently transgressed in the slightest degree.
Areas which are white at birth remain white ever afterward;* areas which

*We leave out of consideration for the present the "peripheral" pigmentation which
albinos as well as spotted animals may possess (see Castle, 105). This is not fully developed
at birth.

IN RATS AND GUINEA-PIGS. 29

are black at birth remain black; those which are yellow remain yellow;
and those which contain at birth black and yellow hairs interspersed remain
in that condition ever afterward. This indicates that every portion of the
epidermis has its pigment-forming capacity early and finally differentiated.
In spotted animals the capacity to form hair and skin pigments is trans-
mitted only to certain portions of the epidermis. Our statistical studies
make it clear that guinea-pigs and rats (and probably other spotted mam-
mals also) transmit with a good deal of constancy definite amounts of
pigmented surface, but that, in guinea-pigs at least, the distribution of this
pigmentation over the body is not strictly localized in the germ. A certain
amount of pigment, apparently, is handed over to the epidermis, but it
seems to be to some extent a matter of chance upon what part of the body
this pigmented epidermis finally comes to lie. It is not, however, entirely a
matter of chance. It is almost certain that the pigment will lie chiefly on
the dorsal surface, and if the pigmentation is not extensive, it will be re-
stricted to one or more of the regions which we have designated eye, ear,
shoulder, side, and rump areas, all of which are paired and frequently separated
one from another by intermediate unpigmented areas. In other cases,
even when adjacent areas are confluent, they show their essential distinct-
ness by sharp differences in color. Besides the paired areas mentioned,
there is an unpaired area at either end of the body. The anterior one we
have designated nose spot, the posterior one might be called a tail spot,
though in the guinea-pig it is scarcely distinguishable from the rump spots,
because there is no external tail. The distinct pigment spots are derivatives,
doubtless, of individual blastomeres set apart early in development for the
production of the epidermis. We know that in birds and mammals the
epidermis is first differentiated along either side of the primitive streak.
The ectoderm along the middle of the primitive streak sinks down to form
the neural canal, then the divided right and left halves of the epidermis
come together above it, while anterior and posterior to the neural invagina-
tion the right and left halves of the epidermis have been from the beginning
continuous. Very likely the nose- and tail-spot regions correspond with
these regions of original continuity of the right and left halves of the epider-
mis, while the paired areas are formed out of the epidermis lateral to the
neural in vagina tion. The epidermis of the ventral side of a bird or mam-
mal is developed, we know, later than the dorsal portions, and in spotted
individuals, apparently, the amount of pigment inherited is insufficient
ordinarily to extend out over the blastoderm into this region. Accordingly
the pigment is restricted to the portions of the epidermis first differentiated,
i. e.} adjacent to the primitive streak. In most cases, it is insufficient to
cover more than portions even of that.

30 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

On this interpretation, the difference between a black-eyed white and a
pink-eyed white (or albino) guinea-pig is this : The neural plate of the former
receives a pigment potentiality from the ectoderm which is invaginated to
form the neural canal (out of which the retina of the eye is later formed) ;
the albino does not receive the pigment potentiality into the neural plate,
simply because that potentiality is absent from the entire ectoderm. In
the black-eyed white guinea-pig no part of the epidermis is pigmented,
because all the color-bearing ectoderm has gone into the neural invagina-
tion and none is left outside for the epidermis. When a very little is left
outside, it usually is found either upon an ear, upon one side of the fore-
head, or about the eye (see Castle 105), facts which indicate with consider-
able clearness what relation the epidermis of the head region bore to the
anterior part of the neural invagination, and support the view that the nose
spot arises out of epidermis originally anterior to the neural plate. Black-
eyed white guinea-pigs usually produce offspring with one or more spots on
the body, *. e., the amount of pigment which they transmit is commonly large
enough to extend lateral or anterior to the neural plate and cause epidermal
spots as well as pigmented eyes. Thus a pair of black-eyed white guinea-
pigs has recently (Nov., 1906) produced one black-eyed white young and two
with nose spots and an ear spot each. In spotted guinea-pigs the pigment
is apparently always centralized about the primitive streak. It invariably
passes into the neural canal (if any pigment* is inherited at all), so that
the eyes are pigmented; usually it extends out also into the epidermis
bordering on the neural plate, but its limits are variable. When the
amount of epidermal pigment is small, it is found most often upon the head,
adjacent to the portion of the neural plate out of which the eyes were
formed.

If we could follow through the period of cleavage the history of the
pigment-forming substance of the fertilized egg, we should probably find
that in spotted guinea-pigs this is distributed to certain blastomeres in
the animal hemisphere of the egg. Apparently it is to some extent a
matter of chance what blastomeres receive the pigment, but the anterior
part of the neural invagination is sure to receive pigment, and the adjacent
portions of the skin are more likely than any other regions to do so. In
nose-spot animals, pigment is evidently distributed to the most anterior
epidermal cells which take part in forming the dorsal surface of the animal.
If, as our observations indicate, the condition is to some extent hereditary,
the pigment potentiality must be localized in both egg and sperm in that

*See footnote, page 28.

IN RATS AND GUINEA-PIGS. 31

part of the germ-cell which gives rise to the anterior part of the primitive
streak. In selecting for the nose-spot character, accordingly, we have
simply selected for extreme anterior localization of a reduced amount of pig-
mentation. Selection for localization less far forward is apparently less
effectual, probably because, in that part of the germ, a cell-division, which
would involve unequal distribution of the pigment potentiality to the
daughter cells, might carry the potentiality either forward or backward in
the embryo, whereas original extreme anterior localization could result in
its transportation only forward to the most anterior part of the embryo.
This, while a purely hypothetical explanation, is offered as a suggestion
to the embryologist, who some day, perhaps, will be able to identify in the
germ and trace through its various stages of development the substance
or substances on the presence of which pigment formation depends.

If anterior localization of pigment is possible through selection, it would
seem that posterior localization should likewise be possible. We began at
one time a series of experiments to test this point, but the results given by
the first selected generation were so unpromising that the experiment was
abandoned. It may be pointed out, however, that one might expect fixa-
tion of such a pattern to be more difficult, because, like the Dutch-marked
pattern, it involves a double selection, viz, selection for the posterior local-
ization of pigment sought (tail spot) and simultaneously and unavoidably
selection for the anterior localization of pigment represented in the black
eyes of all spotted individuals.

The explanation which has been offered for the distribution of pigment in
partial-albino guinea-pigs will apply equally well to the case of rats, with
this difference: In rats the body is more elongated than in the guinea-pig^
and pigment-reduction affects at first rather the lateral than the longitudinal
distribution of pigment. The back-stripe first becomes narrower, then
becomes interrupted, and finally drops out altogether, leaving pigment
only on the head and usually also sparingly on the tail (see pi. i). This
condition corresponds with that condition of the guinea-pig in which head
and tail (or rump) spots only are present, a condition very common in
Series H, as we have seen. As the hood becomes reduced in extent, the
white areas first extend forward ventrally to the mouth, then in the median
dorsal line between the shoulders, and simultaneously a white spot appears
in the forehead (see pi. i, fig. 3). Further reduction than this has not yet
been obtained, but it is evident that the pigmented areas are becoming
restricted toward the sides of the head adjacent to the eyes, precisely as in
guinea-pigs.

32 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

CONCLUSIONS.

The results of selection brought to bear upon the coat-pattern are seem-
ingly very different in rats and in guinea-pigs, yet a careful analysis of the
facts shows the results to be not so dissimilar in the two cases as they at
first thought appear.

In both rats and guinea-pigs we can by selection increase or decrease at
will the average extent of the pigmented areas. In both rats and guinea-
pigs the extent of the pigmented areas varies continuously, and out of these
continuous variations permanent modification of the pigmentation can be
secured.

Reduction in the total amount of the pigmentation is attended in rats
by restriction of the pigment to very definite areas, whereas in guinea-pigs
it may be distributed in any or in all of a series of spots. Herein lies the
whole difference between the two cases. When in rats we select for reduced
pigmentation, we get animals with a narrow or interrupted back-stripe and
with a less extensive hood; when in guinea-pigs we make a similar selection,
we get animals with fewer or less extensive spots. We can not in guinea-
pigs decide arbitrarily which areas shall be pigmented (except, possibly, in
the case of nose spots) , any more than in rats we can at the same time
increase the extent of the hood and decrease that of the back-stripe.

In rats, we have as a result of pigment reduction a series of coat-patterns,
each breeding true within certain limits; in guinea-pigs, the fluctuation in
the extent of the pigmented areas is probably no greater than in rats, but
because the pigmented areas do not disappear in as definite an order during
pigment reduction, we have no constant coat-patterns. Nevertheless there
is every reason to suppose that different degrees of pigmentation are in-
herited in Mendelian fashion in guinea-pigs, precisely as they are in rats.
If the pigment reduction followed a definite course in guinea-pigs, as it does
in rats, this would be easily recognizable in the coat-pattern. As it is,
measurement of the extent of the pigmented areas would be necessary to
make it apparent. This we have not undertaken to do in the case of guinea-
pigs; we have merely taken account of the regions pigmented, not of their
extent. This probably explains in part why regression is observed in the
selection experiments with guinea-pigs, but not with uniformity in those
with rats. In guinea-pigs we attempted by selection to restrict the number
of the pigmented areas ; this was found to be impossible except as it occurred
incidentally to reduction in the total amount of pigmentation. The regres-
sion occurred in number of pigmented areas, not, so far as we know, in the
total amount of the pigmentation. We have no doubt, however, that such
regression would be found to occur in cases in which extreme variates were
selected. We have found it so in selecting black-eyed white guinea-pigs,
those with no pigment except in the eye. Almost invariably the young of

IN RATS AND GUINEA-PIGS. 33

such animals have borne more pigment than did their parents. A similar
result would doubtless follow selection of self-pigmented rats obtained from
Irish parents. No doubt many of the young would bear some white fui.
With selection of less extreme variates, regression less extreme may possibly
occur, though our statistical observations do not show any regression in
the case of rats.*

If regression does occur, can we with propriety consider the effects of selec-
tion permanent? De Vries has answered this question in the negative on
the basis of his selection experiments with maize, striped flowers, double
buttercups, and other similar material. It seems to us, however, that the
answer should be qualified. The final result will depend upon the amount
and the persistency of the regression. In De Vries's experiments with maize,
as in those of Fritz Miiller ('86), the regression grows less with each selection.
If this continued, the regression should ultimately become a negligible quantity.
After repeated selection for a desired extreme condition, the race should
become stable at a condition only a little less extreme than that selected.

De Vries's fine series of selection experiments with the buttercup (Ranun-
culus bulbosus) seems to the writers scarcely to justify the conclusion that
selection has no permanent effects. Starting with a one-sided or "half-
Galton" variation curve, with a range from the modal number, 5, upward
to 13, De Vries was able by selection for an increased number of petals to
raise the mode to u, the average to 8.6, and the upper limit of variation to
31, and to obtain a two-sided, or Galtonian, variation curve with only a
moderate amount of skewness, and with greatly diminished regression. All
this was accomplished within five generations.

We consider the selection question still an open one. Further experiments
and longer continued ones are needed. Our own observations, so far as
they go, and those of Fritz Miiller and De Vries, lead us to think that selection
is a most important factor, not only in the isolation of discontinuous varia-
tions, but also in their production.

Further, we are far from convinced that all evolutionary progress is to be
attributed to discontinuous variations, any more than to Mendelian inher-
itance. The distinction between continuous and discontinuous variations
is a useful one, just as that between alternative and blending inheritance,
but a sharp line of division can be drawn in neither case. The hooded and
Irish coat-patterns of rats are recognized discontinuous variations, alter-
native in inheritance, yet our lot M of hooded rats is as nearly intermediate
between typical hooded and typical Irish rats as anything that can well be
imagined. The coat-patterns of fancy rats, though discontinuous as they
ordinarily occur, can be transformed into continuous variations. Concerning

*April, 1907. In this year's experiments we are getting some evidence of the expected
regression.

34

INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

the hooded and Irish patterns, Doncaster (:o6), after an extended experience,
says (p. 216): "Only once have I had any hesitation in classing a rat as
belonging to one or the other." Yet we have seen that by cross-breeding
and selection these same discontinuous groups can be made continuous by
the production of any desired number of intermediate groups, each varying
continuously about a different mode.

Again, though the inheritance is clearly Mendelian, when hooded and Irish
rats are crossed, the gametes formed by cross-breds are not pure, but modified,
each extracted pattern being changed somewhat in the direction of that
pattern with which it was associated in the cross-bred parent. This means
simply that the inheritance, though in the main alternative, is to some extent
blending.

Since it is impossible to make a sharp distinction between continuous and
discontinuous variations, as well as between blending and alternative inherit-
ance, it is fallacious to assign all evolutionary progress to one sort of variation
or to one sort of inheritance.

TABLES.

TABLE i. — Actually different classes into which we may expect the five visibly different

classes of albino and partial-albino rats to jail.
[Twenty-two of the twenty-seven classes enumerated have been proved to exist.]

Classes visibly different.

Gray Irish.

Gray
hooded.

Black
Irish.

Black
hooded.

Albino.

r

i GI

i GH

i BI

i BH

iW

"BH]

.

2 GKW)

2 GH(W)

2 BI(W)

2 BH(W)

i W

BI]

2 GI(B>

2 GH(B)

2 BI(H)

i W

GH]

Classes
expected

2 GI(H)
4GI(W.B)
4GI(W.H)
4 GI(BII)

4GH(W.B)

4 BIcW.H)

i W

2 W
2 W

2 W[

GIJ
"B-IH]
G-IH]
GB-H]

.

SGI(W.BH)

.

•

2 W[

4w

GB-I]
GB-IH]

Total...

27

9

9

3

16

B, black; G, gray; H, hooded; I, Irish; W, total albinism. Symbols indicating unseen
recessive characters are placed within parentheses ( ), those indicating characters latent
in albinos are placed within brackets [ ].

When two symbols only are used the first refers to color, the second to coat -pattern.

When more than two symbols stand together within brackets, those which refer to
color are placed at the left of a hyphen, those referring to pattern at its right.

Total albinism is indicated by W. When albinism is recessive with other characters
in pigmented individuals, the W will be separated by a period from the symbols designating
the other characters.

The numerals prefixed to the several class-designations indicate the expected frequencies
of the classes when individuals are mated inter se which have the characters indicated by
the designation GI (W.BH).

IN RATS AND GUINEA-PIGS.

35

2. — Expected and observed distribution of young produced by partial-
albino rats mated inter se.
[Abbreviations as in table 1.]

Mating No.

Nature of mating.

Proportions of off-
spring expected on
Mendelian hypothesis.

Observed result
(expected result, italic).

GI

GH

BI.

BH.

W.

GI.

GH.

BL

i

2.4

5
J

6.J

13
15
o

J.*

5
4
o
o

0

0
0

o

0
0
0

o

26

*7.6

5
7

BH.

W.

I

2

3
4
5
6

7
8

9

10

ii

12
13
14

GI(W.BH) bred inter se..
GI(W.BH)xGM(W.BH)
GI(W.BH)XBI(W.BH)..
GI(W.BH)X BH(W)
GI(W.BH)X BH

27
9
9

3

i

i

0
0
0
0
0
0
0

9
9
3
3
i

0

i
3

9
3

i
o

0
0

9
3
9
3
i
i

0

o
o

0
0

9
3

i

i

3
3
3
3

i

0
0

i
3
3

i

3
3
i

16
8
S
4

0
0
0
0

4

2
0

4

2
0

( ii
1 72

1 S
I 9

i 8
1 6.3

i13

( '5
( 6

1 3-*

\\

< 8
\ 9

\ °
1 o

i °

\ o

i °

i 0

( 0

1 o
1 o

1 o

( 0

1 o

! °

i o

2

2.4

10

9

i

2 I

16
'5
3

3-*

0
0
IO

9

18

21

«4

12.4

H
'3-9
4
J-5
o
o

0

o
o
o

0.J

2
J

4

2.1
24

'5
4
,?•*

0

o
o
o

10

7

2

*'
12

'39
3
J5
ii
?.j

20

18
9
7

2
^^

I
5

5-5
H

^0

o
o

0

o

0
0
0

o
6

f-5
ii

92

0

o

12
122

U
12
0
0

GI(W B)X BI(H) . ...

GI(H)XBH(VV)

GH(W B)xGH(B)

GH(W.B) bred inter se...
GH(W B)xBH(W)

GH(W B)x BH

BI(W.H)bred inter se....
BI(W H)X BH(W)

BI(H)XBH

57
5J-6

92

9'- 5

74

86.4

102

S6.8

70
767

INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

3. — Expected and observed distribution of young produced by partial-
albino rats mated uith albinos.
[Abbreviations as in table 1.]

0

tJD

c

H
%

Nature of mating.

Proportions of off-
spring expected on
Mendelian hypothesis.

Observed result
(expected result, italic).

GI.

GH.

BI.

BH.

W.

GI.

GH.

BI.

BH.

W.

15

16

17

18

19

20
21
22
23

GI(W.BH)X\V[BH]

i

0
0

o

0

3
o

0
0

i

i
o

3
o
i

0

o

0

o

3

o

I

i

i
i
i
i
i
o
i

4

2
0
O
0

8

4
i

2

11.

{o
( o

t °

\ °
1 o

\:

\5<,

f o
'( o

is

1°

L

6
II

1J

11.5

0

o

2

*>4
o
o

8

65
o

0

6

8.1
o
o
o

0

o
o

9
//

5
4.1

12
70.5
O
O

24

'7

ii
8.1
6
6
14

12 5

15
77.5

13

77

2

6<

J-5

0
0

'7
'7

34

32.5
H

12
0
O
0

o
o

0

8

77

10
'*

6-5

27

J4

GH(W.B)x W[BH] ...

GH(B XW[BH]

BHxVV[GB-H]

BHxWfB-Iin

BI(W.H)XW[GB-IH] ....
BI(W.H)XW[B-IH]
BH(VV)xW[GHJ

BH(W)XW[B-IH]

12
12.2

40
46

56

507

84
7'

98

770

Totals expected

TABLE 4. — Average grade (size of stripe} in the various lots of hooded rats of selected
parentage, as compared u>ith the average grade of their parents.

Deviation

Lot.

No.

Individuals.

Average
grade.

Average
grade
of parents.

of average
filial from
average pa-

rental grade.

E

132

211

21.8

-0.7

F

89

12.9

17.8

-49

G

83

96

II.O

-i 4

H

61

5-6

7-8

— 2.2

I

34

4.0

+2.3

J

in

15-8

12.0

+3-8

K

70

22.2

23.0

-0.8

L

64

36.2

40.7

-4-5

If

19

524

500

+2-4

IN RATS AND GUINEA-PIGS.

37

5. — Relation between distribution of pigment spots in i ij guinea-pigs of
group D%, and in their parents.

Offspring.

Parents.

Location of
spots and total
number of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 Spots.

i spot.,
o spots.

Total..

2 SpOlS .

i spot.,
o spots.

Total .
2 spots.

4I

I

85.7

12.3

2

10
8

55.6

44-4

35
15

70
30

1 Head ;
^ total cases,

J.,7.

^
1 Shoulder;
}• total cases,
1 "7-

| Side ;
V total cases,
1 "7-

j Rump ;
} total cases,
1 "7-
J

49

...

IS

...

50

4

i

80
20

3
3

2

37-5
37-5
25

8

4
10

36-4

18.2
45-4

...

...

13
21

43

1.59
25.6

53.5

5

...

8

...

22

"...

...

...

82

7

70

3

100

16
10
18

36-4

22.7
40.9

3

i

75
25

'4

3

25
25
50

o spots.
Total..

2 Spots.

i spot..
o spots.

Total. .

3

3»

...

...

10

...

3

...

44

...

4

...

5<>

...

25
6

4

71.4
17.2
II-4

...

...

4i
20

21

5°
24.4
25.6

...

...

...

...

35

...

...

...

82

TABLE 6. — Combination of the four parts of table 5, showing in general the extent to which
pigment spots are inherited in group D±, i ij individuals (468 observations).

Offspring.

Parents.

Total

iiuiiib.-r
of
cases.

U68
J

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

;P. ct.

Cases.

P. ct.

2 SpOtS ..
I Spot ...

o spots...
Total ..

78

78.8
132

So

16

ii

2

55-2
37-9
69

100

49
49

50-5
24.7
24.7

3

...

35

19.6

254

55-o

99

29

...

198

...

4

138 j ...

INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

TABLE 7. — Relation between distribution of pigment spots in 30 guinea-pigs of
group Di, and in their parents.

Offspring.

Parents.

Location of
spots and total
number of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 spots.
I spot. .

o spots.
Total..

2 Spots.

i spot.,
o spots.

Total..

2 Spots.

I spot. .

23

5

82.1
17.9

I

| Head ;
}• total cases,
130.

| Shoulder;
}• total cases,
130.

| Side;
}• total cases,
130.

| Rump;
}• total cases,

130.
J

28

2

...

...

2
2
I

I

4

5
13

18.3

22.7

59-°

...

...

...

I
I

...

...

5

...

...

3

22

...

2
2

...

...

...

13

4
9

50 o
J5-3
347

o spots.
Total..

4

...

...

...

26

...

2 spots.

i spot. .
o spots.

Total..

24
3
3

80

IO
10

30

TABLE 8. — Combination of the four parts of table 7, showing spot inheritance in
group DI, 30 individuals (120 observations).

Offspring.

Parents.

Total
number
of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 Spots...

i spot....
o spots...

Total....

1
3

Si.o
13-8

S-2

3
3

i

...

2
2

...

I
I

...

17
9

22

~

45^8

1

}>I2O
1

J

58

...

7

...

4

...

3

...

...

IN RATS AND GUINEA-PIGS.

39

9. — Relation between distribution of pigment spots in ij8 guinea-pigs of
group D2, and in their parents.

Offspring.

Parents.

Location of
spots and total
numbrr of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 Spots.

i spot..
o spots.

Total .

2 Spots.

I spot..
o spots.

Total .

2 Spots.

i spot.,
o spots.

Total.

2 Spots.

i spot.,
o spots.

Total.

154

24

86.5

13.5

|Head;
^ total cases,
I 178.

1
| Shoulder;
}• total cases,

h

| Side;
}• total cases,
|i78.

| Rump;
V total cases,
(178.

178

...

...

3

i
i

••

5
7
7

26.3
36.8
368

7
4
4

46.7
26.7
26.7

49
33

57

353
23-7
41.

...

5

...

'9

...

15

...

i39

...

6

2

3

54-5
18.2

27-3

4

2

...

22

7

12

53-6
17.1

29-3

5
i
6

41.7

8-3
50-

39

18

5i
108

36-1
16.7
47.2

ii

...

6

...

41

...

12

...

144
16

18

80.9
9.0

10. 1

178

TABLE 10. — Combination of the jour parts of table p, showing spot inheritance in
group D2, 178 individuals (j 12 observations).

Offspring.

Parents.

Total
number
of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 spots...

i spot. ..
o spots...

Total...

304
42
21

82.8
II.4

5-7

7

3

i

63.6

27.3
9.1

27

H

19

45.0
233
31.7

12

5
10

44-4
18.5
37-0

88

&

35-6

20-7

43-7

712

367

...

ii

...

60

...

27

...

247

...

INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

TABLE u. — Relation between distribution of pigment spots in 136 guinea-pigs of
group D3, and in their parents.

Offspring.

Parents.

Location of
spots and total
number of
cases.

4 spots.

3 spots.

2 Spots.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. Ct.

Cases.

P. ct.

2 Spots.
I Spot..

o spots.
Total.

2 Spots.

I spot. .
o spots.

Total.

2 Spots.

I spot.,
o spots.

Total .

2 Spots.

i spot. .
o spots.

Total.

117
H

89.3

10.7

4

I

80

20

| Head ;
)> total cases,

! 136-

| Shoulder;
}• total cases,
1 136.

| Side;
}• total cases,
1 136.

1
1 Rump;
\ total cases.
|i36-

—

131

...

5

..

...

2

...

...

<

...

48

28

55

36.6
21.4
42.0

'

...

...

...

3

2

«3«

...

2
2
2

33-
33-
33-

...

...

38

H

20

52-8
194

27.8

...

...

'9
13
26

32.8

22.4

44.8

6

...

...

...

72

...

...

...

58

...

"3
H
9

83-1
10.3
6.6

136

TABLE 12. — Combination of the four parts of table //, showing spot inheritance in
group D$, 1 36 individuals (544 observations).

Parents.

Total

Offspring.

4 spots.

3 spots.

3 SpOtS.

i spot.

o spots.

number
of

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 Spots...

232

85.0

6

75

38

52.8

,

...

67

35-5

1

i spot....
o spots...

30
II

II. O
4.0

2

25

H
20

19.4

27.8

..!

...

48;

21.7
42.8

!>544

Total....

273

...

8

...

72

...

2

...

i89

...

J

IN RATS AND GUINEA-PIGS. 41

TABLE 13. — Combinations separately of the four parts of tables 5, 7, 9, and //, showing
inheritance of particular spots in series D as a whole, 461 individuals.

Off sping.

Parents.

Location of
spots and total
number of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 Spots.

i spot.,
o spots.

Total..

2 Spots.

I spot.,
o spots.

Total..

2 Spots

i spot. .
o spots.

Total..

2 SpOtS.

I spot. .
o spots.

Total .

336

49

87.0
12.7
o-3

*5
10

60

40

35
15

70

30

| Head;
[•total cases,
1 461.

j Shoulder;
)> total cases,
|46i.

|Side;
}• total cases,
|46i.

| Rump;
}• total cases,
! 461.

J

386

...

25

...

...

4

i

80

20

10

7
4

21

47-6
33-3

KJ.O

13
ii

17

£1

4i-5

9

6

450
25.0
30.0

114
87
173

30-5
23.2

46.3

5

...

...

4i

...

20

374

...

15
8

SS-6
I4.8
29.6

7

2

77-8
22.2

78
33
50

48.4
20.5
41.0

8

2

6

50.0
12-5
37-5

85
49
114

34-3
19.7
46.0

27

...

9

...

161

...

16

...

248

...

306
39
34

80.7
10-3
9.0

...

...

41
20

21

500
24.4
25.6

379

...

...

...

82

TABLE 14. — Combination of tables 6, 8, 10, and 12, showing spot inheritance in
series D in general, 461 individuals (1,844 observations}.

Offspring.

Parents.

Total
number
of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 SpOtS..

i spot....
o spots...

Total...

661
93
43

82.9
11.7

5-4

32
19
4

58.2

345
7-3

167
i

334

50.0

237
26.3

'1

II

47.2
22.2
30.6

199

32.0
21.9
46.1

^844

797

...

55

...

...

36

...

622

...

INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

TABLE 15. — Relation between distribution of pigment spots in 207 guinea-pigs of
group H±, and in their parents.

Offspring.

Parents.

Location of
spots and total
number of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P.ct.

Cases

P. ct.

2 spots.
I Spot..

o spots.
Total .

2 Spots.

i spot. .

8.3

12

9

79.8

"-S

8.7

20
13

33

60.6

394

35
15

2

673

28.8

38

10

3

76.9

23 »

3

2

60.
40.

] Head ;
} total cases,
I 207.

1
| Shoulder;
} total cases,
i 207.

J

1 Side ;
} total cases,

1 207.

| Rump;
V total cases,
1 207.

104

...

...

52

...

13

5

...

H

7

21

33-3
16.7
50.0

...

...

27

£

16.4
194
64.2

o spots.
Total.

2 Spots.

i spot..

42

i6c

20

9

21

40.0
18.0
420

i

5
5

9.1

45-4
45-4

25
3i
90

17.1

21.2
6l.7

o spots.
Total .

2 SpOtS.

i spot . .

50

...

ii

146

...

44
17
23

52-4

2O 2

27-4

10

8
ii

34-5
27.6

379

26
14
54

27.7
I4-9

57-4

o spots.
Total.

84

...

29

...

94

...

TABLE 16. — Combination of the four parts of table 15, showing spot inheritance in
group H\t 207 individuals (828 observations}.

Offspring.

Parents.

Total
number
of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases. 1 P. ct.

Cases.

P.ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P.ct.

2 spots...
I spot

o spots...
Total....

83
12

9

79-8
n-5
8.7

20
13

60.6

39-4

"3

|

49.6
21. 0
29.4

21

16
16

39 6
30.2
30.2

81

77
252

Ill

61.5

1
^828

104

...

33

...

228

...

53

...

410

...

IN RATS AND GUINEA-PIGS.

TABLE 17. — Relation between distribution of pigment spots in 193 guinea-pigs of
group Hi, and in their parents.

43

Offspring.

Parents.

Location of
spots and total
number of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P.ct.

Cases.

P. ct.

2 spots.

I spot. .
o spots.

Total .

2 Spots.

i spot. .

88
II
4

854
10.7

3-9

13

5

68.4
26.3

5-3

37
21

7

57-o

32-3
10.7

3

i

2

6

50.0

16.7

333

...

...

1 Head ;
[-total cases,
1 193-

j Shoulder;
} total cases,
1 193-

Side ;
\ total cases,
193.

| Rump;
V total cases,
i 193-

103

...

19

...

65

...

...

...

...

14
140

39

7.3

20.7
72.0

o spots.
Total.

2 Spots.

i spot..

*93

...

29

35
129

15.0
66.8

o spots.
Total.

2 SpOtS.

i spot..

54

&

28.0
16.0
56.0

o spots.
Total.

;

193

...

1 8. — Combination of the jour parts of table //, showing spot inheritance
in group HI, 193 individuals (772 observations).

Offspring.

Parents.

Total
number
of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 Spots..

i spot. ...
o spots...

Total...

88
ii

4

85.4
10.7

3-9

13

5
i

684

26.3

5-3

37
21

7

56.9
323

10.8

3
i

2

50
16.7

33-3

97
1 06
376

16.8

i8'3

64.9

[772
J

103

19

...

65

...

6

...

579

44

INHERITANCE OK COAT-PIGMENTS AND COAT-PATTERNS

TABI,E 19. — Relation between distribution of pigment spots in 1 29 guinea-pigs of
group 7/2, and in their parents.

Offspring.

Parents.

Location of
spots and total
number of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Ca^es.

P. ct.

Cases

P. ct.

2 spots,
i spot.,
o spots.

Total..

2 SpOtS.

i spot..
o spots.

Total. .

2 SpOtS.

i spot.,
o spots.

Total..

2 Spots.

i spot.,
o spots.

Total . .

74

20

2

96

77.J
20.8
2.1

10

2

83-3
16.7

*7

4

81.0
190

| Head;
}• total cases,
1 129.

| Shoulder;
}• total cases,
1 129.

JSide;
total cases,
129.

1
I Rump;
}• total cases,
1 129.

...

12

...

21

4

99

3-i

20 2

76.7

129

...

ii

3°

88

8-5
23-3

68.2

129

...

36

J2

27.9
58-9

129

...

TAEI,S 20. — Combination of the four parts of table 19, showing spot inheritance in
group //2, 129 individuals (5 / 6 observations) .

Offspring.

Parents.

Total
number
of

cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

P. ct.

P. ct.

2 Spots...

i spot —
o spots...

Total.. .

74
2O

2

771
20.8
2.1

10

2

83-3
16.7

17

4

81.0
19.0

...

...

51

%

13.2

18.8
68.0

^516

96

...

12

...

21

...

...

...

387

...

IN RATS AND GUINEA-PIGS.

TABLE 21. — Relation between distribution of pigment spots in $8 guinea-pigs of
group //3, and in their parents.

45

Offspring.

Parents.

Location of
spots and total
number of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Oses.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 Spots. .
I Spot . .

46
10

82.1
17.9

I
I

1 Head ;
J total cases,
»

Shoulder;
> total cases.
,58-

1 Side;
J total cases,
5,

1 Rump;
> total cases,
,58.

Total..

2 Spots. .

56

...

2

2

9
47

34
'5-5
81.1

i spot . .

o spots..

Total . .

58

...

2 spots..

4
H
40

58

6.9
24.1
69.0

i spot . .
o spots..

Total..

2 Spots .

i spot. .
o spots .

Total . .

...

*3
37

22.4
13.8
63.8

58

...

TABLE 22. — Combination of the four parts of table 21, showing spot inheritance in
group //3, 58 individuals (232 observations).

Offspring.

Parents.

Total
number
of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 Spots...

i spot....

46
10

82.1
17.9

I

19
31
I24

10.9
17.8
71-3

232

Total...

56

...

*

...

174

...

46

INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

TABLE 23. — Combinations separately of the four parts of tables 15, ij, 19, and 21,
showing inheritance of particular spots in series H as a whole, 587 individuals.

Offspring.

Parents.

Location of
spots and total
number of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 Spots.
I SpOt..

o spots.
Total.

2 SpOtS.

291

53
15

81.1
14.7
4.2

44
21

I

66.7
31.8

i-5

89
40

9

64.5
29.0

6-5

13

4

2

68.4

21. 1
10.5

3

2

60.
40.

IHead ;
total cases,
587.

1 Shoulder;
}• total cases,

1587.

ISide;
} total cases,
1 587.

IRump;
total cases,
587.

359

...

66

...

138

...

19

...

5

...

H

7

21

33-3
16.7

50.

...

...

47
107

39i

18.6
19.7
71.7

o spots.
Total .

2 spots.

i spot. .

42

...

...

...

545

...

20

9

21

40.
18.
42.

I

5
5

9.I

45-4
45-4

69
no

347

13-1
20.9
66.0

o spots.
Total.

2 Spots.

50

...

ii

...

526

...

44
17
23

52-4

20.2
27.4

10

8
ii

34-5
27.6

37-9

129

70
275

27.2
14.8
58.0

o spots.
Total .

84

...

29

...

474

...

TABLE 24. — Combination of tables 16, 18, 20, and 22, showing spot inheritance in
series H in general, 587 individuals (2348 observations).

Offspring.

Parents.

Total
number
of
cases.

4 spots.

3 spots.

2 SpOtS.

i spot.

o spots.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

Cases.

P. ct.

2 spots...

i spot —
o spots...

Total...

291
53
15

81.1
14.7

4.2

44
21

I

66.7
31.8

i-5

I67

73

74

53-2
23-3
23-5

24
17

40.7
288
30-5

248
287
1015

16.0
18.5
65-5

1
^348

359

...

66

...

3H

...

59

...

J55°

TABLE 25. — Frequencies with which spots absent in the parents (or in one parent,
in group Hi) are absent also in their offspring.

Group Hi
(table 15).

Group Hi

(table 17).

Group Ha

(table 19).

Group HZ
(table 21).

Average
(table 23).

Shoulder spots
Side spots

TVr ct.
64.2
61.7

Psr ct.
72 o
66.8

Per ct.

76.7
68.2

P<*rct.
8l.I
69.0

Per ct.

71.7
66.0

Rump spots

57-4

56 o

58.9

638

58.0

IN RATS AND GUINEA-PIGS.

47

TABLE 26 — Constants for the four groups of series D, based upon tables 6, 8, /o,

and 12, respectively.

Group D£.

Group Di.

Group Da.

Group Dj.

Average number of spots borne
by a parent . .

3 78

44.O

A 62

A f>±

Average number of spots borne

4-. 76

C.44

T- "f

c.c6

4.LKJ.

5*72

Per cent increase over parents ....
Standard deviation (<y) in num-
ber of spots, parents

26
2.876

24

3 812

20

•5 744

»*

23

7 6o<l

Standard deviation (<y) in num-
ber of spots offspring.

3 CO 7

•j 6ic

•5 663

O UVT
? 662

Coefficient of variability, par-
ents

J^J

76.78

ge 12

80 QO

Coefficient of variability, off-

72.02

67.70

6c.7o

/y T~f

64 20

Coefficient of correlation be-
tween parents and offspring. .

0.4527

0.4580

0.4278

°-4343

TABLE 27. — Constants for tlie four groups of individuals of series Ht based upon
tables 1 6, 18, 20, and 22, respectively.

Group HJ.

Group Hi.

Group H?.

Group H3.

Average number of spots borne
by a parent .

2 4.8

I.C6

I 80

2 OO

Average number of spots borne
by offspring

T 88

•j 20

^oS

•? oo

Per cent increase over parents
Standard deviation (^-) in num-
ber of spots parents

56

2 870

105
'y 022

71
1 ing

50
34.74.

Standard deviation (<r) in num-

•j c76S

•? 6O76

•?.tq6o

•'t^T

3 6200

Coefficient of variability, par-
ents

1IC 72

lS6 C7

177 66

I7l *7O

Coefficient of variability, off-
sprino1...,

Q2 iS

1 12 7C

11^ OO

•/••/**

1 2O 60

Coefficient of correlation be-
tween parents and offspring....

°4565

0.6076

o 6109

06796

TABLE 28. — Character of the young of the original nose-spot male,

(NSo).

Individuals
with
nose spots.

Z
$*

3&

r

Average
number
other spots.

Individuals
without
nose spots.

jj

"* %
1

III
il]

< B-S
0

Per cent of
young with
nose spots.

By mothers without nose spots
By NS0 mothers

4

22

6

58

1-5

2.6

H
19

40

58

2.8

3.0

22.2

S3 6

By NSi mothers

16

•2Q

2.4

io

2Q

2.9

61 «:

?

C

1.6

2

i

o. c

w«O

60.0

Total

4C

108

2.4

4C

128

2.8

CO.O

48 INHERITANCE OF COAT-PIGMENTS AND COAT-PATTERNS

29 — Cliaracter of the young of nose-spot males, chiefly descendants of 19890 (table 28},

•3 §

1

«

•3« a

i

i

M'j

-ajs a,

o *

rt*J) §,

|||

o 15

1t£ §•

c ^ §.

'>•! u

li

|||

•38.

> 1 S

" c"w

1 I

0

^•5 s

0 *

•< c-g

*" 3 o

H

<:C|

5 §

H

o

(2 >^s

c?5595« NSo by NS0 mothers

x

5

5-

6

27

45

H-3

NSi
6^5678, NSi NSo

4

0

10

o

o.

5

2

15

5

3-

25

44-4
o.

NSX

2

6

3-

3

8

2-7

40.

6^5151, NS2 NS0

15

30

2.

22

66

3-

4°-5

44 44 NSi

9

32

3-5

13

52

4-

4° 9

^5652, NS2 NS0

7

16

2-3

5

13

2.6

583

NS2

2

3

2

9

4.5

c?5388, NS,*, NSo

0

o

0.

4

14

3-5

0.

NSi

5

6

1.2

6

ii

1.8

45-4

(^5669, NS2*, NSo

3

3-

4

10

2-5

20.

44 , NSz

2

5

2-5

i

i

i.

67.

" " , NS2

I

5

5-

i

4

4-

50-

Total

40

121

2.5

74

275

1.2

39.8

1^ 7

O

/ T

oo

•j

TABLE 30. — Further analysis of the data contained in table 29.

•3 a

11!

it

«*.£ £•
Sfltt

> 3 £

|if

-£ .

«"B *

> 3 5

lit
"?J

^ ? o

*^ P rG

V ^ O

i— i C

H

0

>S c

H

O

».Ra

NSo males by NS0 females

I

5

5-°

8

32

40

ii

NSo males by NSi females

6

16

2.7

8

23

2-9

43

NSi or 1^ males by NSo females.

23

49

2.1

35

103

2.9

396

NSi or,*malesbyNSi females.

16

43

2.7

20

64

3-2

44

NSi or i^. males by NSi females.

3

8

2 7

3

13

4-3

50

Totals for NS0 males

7

21

3.0

16

55

34

3° 4

Totals for NSi or Tj males...

42

100

2.2

58

180

42

Totals for NS0 females

24

54

2.1

43

IIS

- i

35-8

Totals for NSi females

22

59

2 7

28

87

3.1

44

Totals for NS2 females

3

8

27

3

13

43

50

Grand totals

4O

121

2.5

74

275

7.2

39-8

IN RATS AND GUINEA-PIGS.

49

31. — Character as regards nose spots of the young of animals of Series H
(head spot series). None of the parents bore nose spots.

Father.

°f
H

Spots of
mother
(average).

Spots, of
parents
(average).

Individuals
with
nose spots.

$S

r

Average
number
other spots.

Individuals
without
nose-spots.

Total other
spots.

Average
number
other spots.

Per cent of
young with
nose spots.

1716*

I

2.O

•5

4

16

4.0

29

100

3-4

12. I

2537

0

2-5

•*5

10

19

1.9

63

197

3-1

38.7

2625
438S

I

0

1.9

3-2

:i

H

4

32

7

2.3
i-7

47
34

132
92

2.8
2.8

23
10.5

4399

I

2.7

.8

7

M

2

21

48

2.3

25

5420

I

i-7

•3

...

...

6

18

3-o

O

Total .

0.56

2-3

14

39

88

2.3

200

589

2.9

16.3

*By an NS0 female, this same male (1716) had ten young, three of which bore nose spots. The average number
of spots other than nose spots borne by ihe NS young was 2.6; by the remaining young, 3.7.

Data for curves (figs. 2 to 5).

Classes...

4-5

'4-5

24-5

34-5

44-5

5^-J

64.5

74-5

Average.

A

h r

64.

•tc

11 "t

B

2Q

7C

7,4

1C

6

7

A

L6-6
21.2

c

8

7

9

2

i

27 C

D

4?

C7

2Q

9

IT

17

I I

i

27 Q

E
p

24

•32

48

42

29

12

16

IO

5

21. 1

T-7 Q

** y

Classes...

2

7

12

'7

22

*7

^

...

...

G

28

17

\n

12

6

2

I

Q 6c

1{

•5 I

1 /
2O

7

2

I

y wj

r A

I

Ol

24.

6

Classes...

4-5

<4-5

24-5

345

^•5

54-5

645

7*. 5

...

7

74.

A A

2 1

10

j

i

1C 76

£:..:.:"

8

TT

18

16

i

I

22 2

L
M

4

9

'4
i

10

2

9

11
6

4

c

3

36.2
C2.4

BIBLIOGRAPHY.

, G. M.
: 04. The heredity of coat-color in mice. Proc. Amer. Acad. Arts and Sci., vol. 40,

No. 2, pp. 61-163, 2 figs.
BATESON, W.

'94. Materials for the study of variation, treated with especial regard to discon-
tinuity in the origin of species, xv+598 pp. Macmillan & Co., London
and New York.
BATESON, W.

: 03. The present state of knowledge of colour-heredity in mice and rats. Proc.

Zool. Soc. London, 1903, vol. 2, pp. 71-93.
CASTLE, W. E.

: 05. Heredity of coat-characters in guinea-pigs and rabbits. Carnegie Institution

Publication No. 23, 78 pp., 6 pi.
CASTLE, W. E.

: 06. The origin of a polydactylous race of guinea-pigs. Carnegie Institution Pub-
lication No. 49, pp. 15-29.
CASTLE, W. E., and ALLEN, G. M.

103. The heredity of albinism. Proc. Amer. Acad. Arts and Sci., vol. 38, No. 21,

pp. 603-622.
CASTLE, W. E., and FORBES, A.

: 06. Heredity of hair-length in guinea-pigs and its bearing on the theory of pure

gametes. Carnegie Institution Publication No. 49, pp. 3-14.
CORRENS, C.

:oi. Die Ergebnisse der neusten Bastardforschungen fur die Vererbungslehre.

Bericht der Deutsch. Botanischen Gesellschaft, Bd. 19, pp. 71-94.
CRAMPE.

'77- 84. Kreuzungen zwischen Wanderratten verschiedener Farbe. Landwirths.

Jahrb., Bd. 6, p. 384; Bd. 12, p. 389; Bd. 13, p. 692.

'85. Die Gesetze der Vererbung der Farbe. Landwirths. Jahrb., Bd. 14, p. 539.
CUENOT, L.

: 02. La loi de Mendel et rh£r£dit£ de la pigmentation chez les souris. Arch. Zool.

Exper. et G6n., S£r. 3, Tom. 10, Notes et Revue, pp. xxvii-xxx.
DARBISHIRE, A. D.

: 03. Second report on the result of crossing Japanese waltzing mice with European

albino races. Biometrika, vol. 2, pp. 165-173, figs. 1-6.
DONCASTER, L.

:o6. On the inheritance of coat-colour in rats. Proc. Cambridge Phil. Soc., vol. 13,

pp. 215-227.
HURST, C. C.

: 05. Experimental studies on heredity in rabbits. Jour. Linnean Society, vol. 29,

pp. 283-324.
LOEB, LEO.

'97. Ueber Transplantation von weisser Haut auf einem Defect in schwarzer Haut
und umgekehrt am Ohr des Meerschweinchens. Arch. Entw.-Mech., Bd. 6,
PP. i-44, 3 Taf., 2 Fig.
MULLER, FRITZ.

'86. Ein Ziichtungsversuch an Mais. Kosmos, Bd. 19, pp. 22-26

VRIES, H. DE.

: 01-03. Die Mutationstheorie. Versuche und Beobachtungen iiber die Entstehung
der Arten im Pflanzenreich. Bd. i, xii-f648 pp., 8 Taf.; Bd. 2, xiv-f 752
pp., 4 Taf. Veit & Co., Leipzig.
103. Species and varieties; their origin by mutation, xviii+847 pp. The Open

Court Pub. Co., Chicago.
WOODS, F. A

103. Mendel's laws and some records in rabbit breeding. Biometrika, vol. 2, pp.
299-306.

50

28846

THE UNIVERSITY LIBRARY

UNIVERSITY OF CALIFORNIA, SANTA CRUZ

SCIENCE LIBRARY

This book is due on the last DATE stamped below.

50m-4,'69(J7948s8)2477

3 2106 00252 1992